David Sobel, Peter Vallentyne, Steven Wall - Oxford Studies in Political Philosophy, Volume 1-Oxford University Press (2015) PDF

OUP CORRECTED PROOF – FINAL, 25/03/15, SPi
OXFORD STUDIES IN POLITICAL PHILOSOPHY

Oxford Studies in
Political
Philosophy
Volume 1
Edited by
D AV I D S O B E L , P E T E R VA L L E N T Y N E ,
A N D S T E V E N WA L L
1
1
Great Clarendon Street, Oxford, OX2 6DP,
United Kingdom
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide. Oxford is a registered trade mark of
Oxford University Press in the UK and in certain other countries
© the several contributors 2015
The moral rights of the authors have been asserted
First Edition published in 2015
Impression: 1
All rights reserved. No part of this publication may be reproduced, stored in
a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by licence or under terms agreed with the appropriate reprographics
rights organization. Enquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above
You must not circulate this work in any other form
and you must impose this same condition on any acquirer
Published in the United States of America by Oxford University Press
198 Madison Avenue, New York, NY 10016, United States of America
British Library Cataloguing in Publication Data
Data available
Library of Congress Control Number: 2014953972
ISBN 978–0–19–966953–0 (Hbk.)
ISBN 978–0–19–966954–7 (Pbk.)
Printed and bound by
CPI Group (UK) Ltd, Croydon, CR0 4YY
Links to third party websites are provided by Oxford in good faith and
for information only. Oxford disclaims any responsibility for the materials
contained in any third party website referenced in this work.
Acknowledgments
Most of the chapters assembled here were first presented at a workshop in

Tucson, AZ in October 2013. We would like to thank all those who attended
this event, and especially the graduate students in philosophy at the University
of Arizona, as well as Rosie Johnson and Gayle Siegel, who assisted in its
organization. David Schmidtz and the Center for the Philosophy of Freedom
provided generous financial support for the workshop, for which we are
grateful. All of the chapters in this volume were reviewed by referees, most of
who serve on the editorial board of Oxford Studies in Political Philosophy. We
very much thank these referees for their efforts in helping to make this first
volume a success. Finally, we would like to express our gratitude to Peter
Momtchiloff for supporting this new series, and for his expert guidance and
advice as we brought this first volume to completion.
Contents
List of Contributors ix
Introduction 1
Steven Wall
Part I: Democracy

1 Justice: Social and Political 9
Philip Pettit
2 Voting and Causal Responsibility 36
Geoffrey Brennan and Geoffrey Sayre-McCord
Part II: Political Liberalism and Public Reason

3 Political Liberalism: Its Motivations and Goals 63
Charles Larmore
4 Political Liberalism, Political Independence, and
Moral Authority 89
Dale Dorsey
5 Against Public Reason 112
David Enoch
Part III: Rights and Duties

6 Territorial Rights: Justificatory Strategies 145
A. John Simmons
7 Can Reductive Individualists Allow Defense Against
Political Aggression? 173
Helen Frowe
8 Elbow Room for Rights 194
Eric Mack
9 Rules and Rights 222
Jonathan Quong and Rebecca Stone
10 What Is Wrongful Exploitation? 250
Thomas Christiano
viii Contents
Part IV: Method

11 Value-freeness and Value-neutrality in the Analysis
of Political Concepts 279
Ian Carter
Index 307
List of Contributors
Geoffrey Brennan is Distinguished Research Professor in the Department

of Philosophy at the University of North Carolina at Chapel Hill and
Professor at the Research School of Social Sciences at the Australian
National University.
Ian Carter is Associate Professor in Political Philosophy in the
Department of Political and Social Studies at the University of Pavia, Italy.
Thomas Christiano is Professor of Philosophy and Law at the University
of Arizona.
Dale Dorsey is Associate Professor of Philosophy at the University of
Kansas.
David Enoch is Professor of Philosophy and Jacob I. Berman Professor of
Law at the Hebrew University of Jerusalem.
Helen Frowe is Wallenberg Academy Research Fellow in the Department
of Philosophy at Stockholm University, where she directs the Stockholm
Centre for the Ethics of War and Peace.
Charles Larmore is W. Duncan MacMillan Family Professor of the
Humanities and Professor of Philosophy at Brown University.
Eric Mack is Professor of Philosophy at the University of Tulane.
Philip Pettit is Laurance S. Rockefeller University Professor of Politics
and Human Values at Princeton University.
Jonathan Quong is Associate Professor of Philosophy at the University of
Southern California.
Geoffrey Sayre-McCord is the Morehead-Cain Alumni Distinguished
Professor in the Department of Philosophy at the University of North
Carolina at Chapel Hill, where he is also the Director of the Philosophy,
Politics and Economics Program.
A. John Simmons is Commonwealth Professor of Philosophy and
Professor of Law at the University of Virginia.
Rebecca Stone is Assistant Professor of Law at the University of California
at Los Angeles.
Introduction
Steven Wall
With this inaugural volume, we are pleased to introduce Oxford Studies in

Political Philosophy, the newest entry in the Oxford Studies series. Since
its revival in the 1970s, political philosophy has been a vibrant field in
philosophy—one that intersects with jurisprudence, normative econom-
ics, political theory in political science departments, and just war theory.
In this and subsequent volumes, we aim to publish some of the best con-
temporary work in political philosophy and these closely related subfields.
This first volume features eleven chapters, most of which were presented
at a workshop in Tucson, AZ in October 2013. The chapters address a
range of central topics and represent cutting-edge work in the field. They
are grouped into four themes: democracy, political liberalism and public
reason, rights and duties, and method.
The first two chapters take up issues in democratic theory. Philip Pettit’s
“Justice: Social and Political” distinguishes social from political justice and
argues that, given some fairly basic assumptions, the latter should take pri-
ority over the former. (Social justice depends on the horizontal relationships
between the individual citizens of a state, whereas political justice depends
on their vertical relationships to the state that rules over them.) Pettit con-
tends that political justice requires a process of decision-making for select-
ing social justice proposals that secures both democratic approval and
democratic control. Democratic approval is necessary since all plausible
theories of social justice advance public recommendations that purport to
address every citizen of the society. Democratic control is necessary because
political justice is a modally demanding good in the sense that its realization
should not depend on the good will of others and because only recognizably
democratic processes of decision-making can plausibly satisfy this demand.
Democratic procedures give each able-minded adult citizen a voice in the
politics of their society. But many are skeptical about the efficacy of this
voice. Can it be rational for someone to vote in a democratic election on
the grounds that he or she intends to make a difference to the outcome?
2 Steven Wall
Standard rational choice theory gives a negative answer. The probability

that one’s vote will make a difference to the outcome is so low that a rational
voter will not include that prospect as being a significant reason to vote.
But, as Geoffrey Brennan and Geoffrey Sayre-McCord point out in “Voting
and Causal Responsibility,” this negative answer hinges on an assumption
about causal efficacy that has been effectively challenged by Alvin Goldman
and Richard Tuck. The assumption is that causal efficacy requires that, if
one had voted differently, the outcome would have been different. Against
this, Goldman and Tuck contend that a vote is causally efficacious just in
case it is among the votes that are, in a suitable way, sufficient for the out-
come. While sympathetic to this critique of the standard rational choice
approach to voting, Brennan and Sayre-McCord go on to present a probing
discussion of the Goldman/Tuck account. They find reasons to doubt both
its explanatory and normative aspirations.
Whether or not it is rational for citizens to view their individual votes
as causally efficacious, democratic processes of decision-making are widely
accepted as a fair way to resolve political disagreements. But many philoso-
phers contend that these procedures, if they are to yield legitimate political
decisions, must be supplemented by an account of the type of reasons that
appropriately can be appealed to in justifying decisions in a pluralistic society.
Following the lead of John Rawls, this approach to democratic legitimacy
introduces the complementary, and now widely influential, ideas of political
liberalism and public reason. The second set of chapters in this volume crit-
ically discusses these ideas. Charles Larmore’s “Political Liberalism: Its
Motivations and Goals” aims to cast light on political liberalism by clarifying
the problem that gave rise to it, the moral foundation on which it rests, and
its limitations. Political liberalism is a distinctive form of liberalism, but it is,
Larmore points out, a latecomer in the history of liberal political thought. It
arose as a response to perceived weaknesses in older versions of liberalism. As
Larmore sees it, political liberalism was a response to the growing recogni-
tion of the significance of reasonable disagreement, including reasonable dis-
agreement over liberal ideals, in modern societies. The political liberal
response to reasonable disagreement appeals to the moral principle of respect
for persons. This principle directs us to seek principles for regulating our
political life that could be the object of reasonable agreement. However,
Larmore stresses that this principle itself need not be the object of agreement,
and reasonable people—that is, people who exercise their general capacity of
reason in good faith and to the best of their abilities—may not accept it. Like
other political conceptions, political liberalism is, therefore, exclusionary. It
excludes those who do not accept its foundational ideas. But Larmore insists
that justification for the political liberal must come to an end somewhere.
Introduction 3
While potentially controversial, the authority of the moral principle of

respect for persons must speak for itself.
In their respective contributions, Dale Dorsey and David Enoch present
spirited challenges to political liberalism. Political liberalism directs us to
“bracket” controversial views of the good when attempting to justify politi-
cal proposals to our fellow citizens. Dale Dorsey takes aim at this feature of
the view in his chapter “Political Liberalism, Political Independence, and
Moral Authority.” He argues that, if political liberalism is accepted, then
our political and moral obligations can come into conflict in a troubling
way. This follows from a basic fact about morality. Morality, unlike political
justice as the political liberal conceives it, does not direct us to set aside our
views about the good. As a result, under political liberalism, someone can
be morally required to do something and required by political justice not to
do it. Dorsey considers a range of responses available to the political liberal
for negotiating this tension, but argues that all of them bring substantial
costs.
David Enoch’s “Against Public Reason” contests the ideal of public reason
that is central to political liberalism and related views. He argues that all
public reason accounts must idealize the subjects of justification and that
this idealization, on inspection, is vulnerable to powerful criticisms. He
argues further that public reason theorists misconstrue the relevant consid-
erations that citizens act on when they impose on others principles that
they, but not these others, accept. The content of the considerations need
make no reference to the fact they are accepted by those advancing them.
Enoch concludes by arguing that the best reason to support public reason
restrictions can be accommodated while rejecting the conclusions stand-
ardly associated with public reason views.
The next set of chapters concerns the rights and duties of persons and the
states that claim to represent them. Modern states under international law
are territorial entities. They claim, and are widely viewed, as having both
legal and moral rights to the territories over which they claim jurisdiction.
Both A. J. Simmons and Helen Frowe address problems raised by the terri-
torial claims that states make. In “Territorial Rights: Justificatory Strategies,”
Simmons identifies a number of strategies for justifying state claims to exer-
cise control over their particular territories. The strategies include volunta-
rist, functionalist, and nationalist approaches to the problem. He argues
that none of these strategies succeeds in vindicating the territorial rights
that actual states claim. A good account of territorial rights then must iden-
tify the kinds of wrongs that states now commit and suggest ways by which
these wrongs can be made right. Simmons argues that a Lockean voluntarist
approach is superior to rival approaches on this account.
4 Steven Wall
Suppose now that an actual state does have justified claims to govern in
the particular territory in which its members reside. Could it defend those
claims against unjustified aggression? Most people think so; but Frowe points
out in her “Can Reductive Individualists Allow Defense Against Political
Aggression?” that explaining why this is the case presents a challenge to any-
one who accepts a reductive individualist view of the morality of war.
(Reductive individualists hold that the rules governing killing in war are
equivalent to the rules that govern killing between individual people.) In
particular, Frowe discusses the argument of David Rodin that states do not
have the right to use lethal force to defend themselves against the unjust
aggression of another state if this aggressor state threatens only non-vital state
interests, such as the interests the state has in maintaining its jurisdiction
over its territory. An aggressor state could claim that its invasion will become
violent only if the victim state resists. The goods threatened by this type of
conditional force aggression, Rodin claims, are not weighty enough to justify
lethal self-defense by the victim state, at least on a reductive individualist
view. Frowe’s chapter presents a substantial response to Rodin’s challenge.
Among other things, she argues that non-vital interests, such as those impli-
cated in political rights to control a territory, when aggregated across enough
people do become weighty enough to justify proportionate lethal self-defense
and that once these rights are violated the threat to the victims changes sig-
nificantly, thus further justifying violence to defend them.
Next up is Eric Mack’s “Elbow Room for Rights,” which tackles a diffi-
cult problem for those who affirm a strong account of self-ownership. Self-
ownership rights theorists, such as Robert Nozick and Mack himself, hold
that people have very stringent rights over themselves and their legitimately
acquired extra-personal property. The problem with this view—a problem
pressed skillfully by Peter Railton and David Sobel in separate articles—is
that these rights appear to be so stringent that they implausibly restrict their
own exercise. For example, the rights seem to rule out very minor noncon-
sensual physical intrusions upon other right-holders.
But, if so, then right-holders may be “morally hog-tied” by the rights of
others, effectively disabled from exercising their own rights because doing
so would violate the rights of others. Mack agrees that this result is unac-
ceptable for the self-ownership view. In response, he presents an explanation
for the permissibility of minor physical intrusions by appealing to the idea
that the specification of the claim-rights of rights-holders must not system-
atically prevent them from exercising the liberty-rights that the claim-rights
purportedly protect. This “elbow room postulate,” Mack contends, allows
the self-ownership theorist to respond to the problem of minor intrusions
without downgrading the stringency of the rights that are central to the
self-ownership view.
Introduction 5
Mack seeks to identify the rights, and limits to those rights, that people
have under favorable conditions. But might our rights change when cir-
cumstances are less favorable? This important and neglected issue is taken
up in Jonathan Quong and Rebecca Stone’s “Rules and Rights.” They pres-
ent a deontological approach to defining rights under non-ideal condi-
tions, one that differs from both remedial and rights-consequentialist
approaches to the problem. The key to their approach is to conceive of
rights violations as a cost or burden that should be distributed fairly among
free and equal persons. Rules that address this distributive problem can
ground individual rights that differ in content from the rights that people
would have under better conditions. In this way, Quong and Stone argue,
our rights change under non-ideal conditions, but they continue to reflect
the underlying values of freedom and equality that underwrite our rights
under ideal conditions.
While non-ideal conditions may affect the rights we have, they cer-
tainly create circumstances in which some people can take unfair advan-
tage of others. Thomas Christiano’s “What Is Wrongful Exploitation?”
presents a general account of unfair advantage taking. He argues that
wrongful exploitation, while wrong, is not a basic wrong. It piggybacks on
other wrongs. In particular, wrongful exploitation occurs when the
exploiter benefits at the expense of the exploited party by violating a duty
he owes to her. Christiano distinguishes his account from other accounts
that have been proposed and illustrates it with a number of examples. The
examples aim to bring out the key features of exploitative interaction, and
show that while it is parasitic on other wrongs it remains a distinctive
kind of wrong itself.
The final chapter by Ian Carter engages with fundamental issues of
method in political philosophy. “Value-freeness and Value-neutrality in the
Analysis of Political Concepts” asks what role does ethical evaluation play in
the analysis of central political concepts, such as power, freedom, and
democracy. Carter’s answer is subtle. He distinguishes value-freeness from
value-neutrality. The former refers to the absence of evaluative terms in the
definition of a political concept while the latter requires that one suspend
judgment on the merits of competing ethical positions when articulating a
political concept. (Neither of these ideas implies that one must detach an
analysis of a political concept from ethical concerns altogether.) Once prop-
erly distinguished, the notions of value-freeness and value-neutrality, Carter
argues, can both inform conceptual analysis of political concepts and help
to clarify our political ideas. In particular, value-free analyses are needed to
relate political concepts to the empirical world; and value-neutral analyses
are helpful in constructing normative theories on which different parties can
converge, allowing them to accept specific evaluations and prescriptions,
6 Steven Wall
despite their disagreements. Ultimately, Carter concludes, it is the failure to

distinguish the different ways by which ethical evaluations can bear on
political concepts that has led political philosophers to misunderstand the
purposes that value-free and value-neutral analyses of political concepts can
serve in our theorizing about politics.
PA RT I
DEMOCRACY
1
Justice: Social and Political
Philip Pettit
1. THE ISSUE
Every year, a number of different agencies rank the world’s cities in order of
livability, taking account of a variety of factors that affect people’s living
conditions.*1 The factors that bulk largest in the ratings are safety and crime,
urban design and environmental quality, public transportation, cultural
tolerance, business conditions, medical care, and so on. These elements all
belong in the domain of what government can provide or facilitate for
people but hardly bear at all on the democratic quality of the government
that ensures their delivery. Thus, the rankings treat people as relatively
passive beneficiaries of urban amenities and ignore the extent to which they
may act as active citizens to maintain control over what government does—
the extent to which they may combine to ensure that government operates
on terms that they generally endorse.
This imbalance of perspective in the international rankings of cities is
intelligible in light of what is allegedly one of the principal aims of the exer-
cise: to provide employers with information on how far employees can claim
hardship allowances for job relocation. Foreign employees need not have a
personal interest in the basis on which government provision for urban resi-
dents is determined, since they are not sufficiently permanent as residents to
be able to claim the democratic rights of citizens. Or at least they need not
have a personal interest in the mode of government decision-making so long
* I benefited enormously from the discussion of versions of this chapter presented

at events in the University of Pavia; the University of Arizona, Tucson; the University
of New South Wales, Sydney; Ohio State University; and as the 2013 David Norton
Memorial Lecture at the University of Delaware. I was helped in earlier stages by comments
from Ian Carter, Emanuela Ceva, Tom Christiano, Dave Estlund, Geoff Sayre-McCord,
Nic Southwood, and Laura Valentini. And I was very helpfully guided in finalizing the
chapter by comments from an anonymous referee and especially by comments received
from Richard Arneson and Steven Wall.
10 Philip Pettit
as the prospect of their being treated well remains bright or they retain a
ready right of exit in the event of that prospect’s fading.
When we think about the different ways in which our society may be
organized—the different institutional structures that may be established
there—we focus on the rival attractions of competing basic structures in a
way that parallels the focus on the rival attractions of the world’s cities. But in
this exercise it is vitally important that we look at those basic structures from
the perspective of active citizens as well as from the perspective of relatively
passive beneficiaries. It would be a serious mistake to enthuse about the way
our needs are satisfied under a certain institutional structure—say, a structure
in which a benevolent despot looks after our needs—without worrying about
how far we are in a position to combine with others to guard against any shift
in the will of the powerful, dictating the terms on which government operates.
This observation teaches an important lesson about what we should take
into account in thinking about the justice of a basic structure.1 In determining
whether a structure is just we have to think about whether it treats the people
who live under it—in particular, adult, able-minded, more or less permanent
residents2—as equals in what we take to be relevant respects; if we focused on
some other property of the structure we would not be thinking about its justice.
Thus, we have to take an impartial viewpoint that is capable of being shared
among the adult, able-minded, more or less permanent members of the
society, or at least among those who are ready to live on equal terms with
others (Pettit 2012; 2014).
The lesson of our observation in the case of livable cities is that in thinking
about the justice promised by a basic structure we ought to give consideration
to two aspects of the structure: first, the social justice between different indi-
viduals and groups that it would establish and, second, the political justice of
the relationship that it would set up between those individuals—roughly,
the citizenry—and the state or government that implements it. Social jus-
tice, as I shall use the phrase, is something that the state provides for its
people as the more or less passive beneficiaries of the system. Political justice
is something that it delivers for its people in their role as active citizens: that
is, as members of the society who ought presumptively to share in setting
the terms on which their state acts. In what follows, I shall assume that the
citizenry are identified appropriately and that they are the adult, able-
minded, more or less permanent residents of the society.
1
I ignore here the issue as to whether justice is properly restricted to the basic struc-
ture of a society without attention to the justice exhibited by individuals acting under
that structure. For a critique of the restriction, see Cohen (1989).
2
Justice imposes special requirements on how those who are not adult, not able-
minded, or not permanent residents should be treated but I ignore those in the present
context. I am concerned with general issues of domestic justice as distinct from special
issues of domestic justice or, indeed, issues of international justice.
Justice: Social and Political 11
Whether a given state is politically just depends, under a broad conception,

on two issues: (i) whether it is just in general that people should have to live
under state coercion; and (ii) whether that particular state exercises coercion
over its citizens on a just basis.3 Here, I consider political justice only under
a narrow conception that sets aside the issue raised by the existence of state
power and focuses exclusively on the issue of how that power is exercised.
Where social justice depends on the horizontal relationships between the
individual citizens of a state, whether individually or in groups, political
justice in this sense depends on their vertical relationships to the state that
rules over them; it turns on how well the state respects their presumptive
claims, collective and individual, to have a say in how it operates. The two
ideals are distinct although, as we shall see in the final section, they have
close connections with one another.
It would be a mistake in the theory of justice to concentrate on social
justice without regard for political, or indeed to pay attention to political
justice and neglect social. Yet, many theories of social justice—often cast
as theories of justice, period—give little or no attention to the political
dimension. And many theories of political justice—in effect, many theories
of democracy—ignore issues of social justice.4
John Rawls (1971) appears to put both on the table when he argues that
apart from guaranteeing standard liberties, and a variant on socio-economic
equality, the just basic structure ought to establish certain political liberties
such as the liberty to vote and stand for office. But he may downgrade polit-
ical as distinct from social justice when he goes on to describe these political
liberties as “subordinate” to the more standard liberties of speech, associa-
tion, and the like (233).5
Assuming that social and political issues are both important in the theory
of justice, I look in this chapter at how they ought to be weighted in relation
3
I consider political justice under a broad conception of what it involves in Pettit
(2012; 2014). Under this conception it coincides with what is often described as the
legitimacy as distinct from the justice—I would say, the social justice—of the structure;
see Simmons (1999).
4
There are exceptions, of course. Thus, Juergen Habermas (1995) has always insisted
on their connection as does Rainer Forst (2002), who operates within broadly the same
tradition. Charles Beitz (1989) and Thomas Christiano (2008) both focus on the impor-
tance of equality in democracy and as a result they also tend to maintain a connection
between concerns of social and political justice.
5
But while Rawls (1971) suggests that the political liberties are subordinate in an
instrumental role, he stresses their intrinsic importance in shaping the character of citi-
zens. And in his later reply to Habermas, Rawls (1995, 163) goes further still in their
defense, implying in the terminology of his interlocutor that the political liberties are
“co-original and of equal weight” with other basic liberties. I am grateful to Paul Patton
and Steven Wall for alerting me to these aspects of Rawls’s position. See too n 9. For a
congenial assessment of how far Rawls can take this later line and preserve his theory of
justice as fairness, see Wall (2006).
12 Philip Pettit
to one another. It should be clear that, like anyone else’s, your ideal of polit-
ical justice—say, your rich democratic ideal of political justice—is liable,
should it be realized, not to support your ideal of social justice: other people
may not combine democratically to support it. And that possibility raises
some serious questions.
Should you be prepared to pay a cost in one form of justice for a benefit
in the other? And if so, how should the trade-off go? Should you be ready,
at the limit, to drop all demands of political justice for the realization of
social? Or should you be ready to drop all demands of social justice for the
realization of political?6
I argue that under some fairly standard assumptions there is a powerful
case for giving a certain priority to political. The very activity of proposing
an ideal or theory of social justice presupposes under those assumptions that
there ought to be a politically just, suitably democratic procedure in place
in the society targeted. And while the content of an ideal or theory of polit-
ical justice may require a degree of social justice in the targeted society—we
shall see in the conclusion that any plausible theory will require this—the
activity of proposing such an ideal is not tied in the same way to a presup-
position of social justice.
In the following three sections I defend three progressively stronger the-
ses about the priority that political justice ought to enjoy under standard
assumptions and then I conclude in section 5 with a brief comment on the
theoretical and practical implications. The theses that I defend in the fol-
lowing three sections hold that as theorists of social justice we are commit-
ted by assumptions almost all of us share to holding that in any society for
which we prescribe a scheme of social justice:
• some politically just process of decision-making ought to be established;
• this process ought to require a form of democratic approval; and
• this process ought to introduce a form of democratic control.7
6
While these issues are often neglected among political philosophers, Laura Valentini
(2013) is a notable exception. I find her views broadly congenial and I was stimulated in
my own thinking about the issues by a presentation that she made in a seminar at
Princeton in a debate with Ryan Davis.
7
The argument of the chapter takes us from assumptions we allegedly share as theo-
rists of social justice to commitments that they entail about the importance and, indeed,
priority of political justice. The argument skirts issues of feasibility, since it explores a
connection between our views of social justice and our views of political rather than
directly supporting any particular institutional recommendations. But it raises a related
question as to what we ought to think if we judge that the commitments in political
justice that we make as theorists of social justice are so demanding as to be unrealizable.
The lesson would seem to be that we should give up on some of the starting assumptions
about social justice but I do not explore that issue in the chapter.
2. SOME POLITICALLY JUST PROCESS OF

DECISION-MAKING OUGHT TO BE ESTABLISHED
Few if any contemporary thinkers who take social justice seriously hold by
their particular view of justice in the manner in which the protagonists in
the European wars of religion held by their different positions. However
fully and fervently they believe in their particular vision of social justice,
they shrink from maintaining that it ought to be imposed by the coercive
state without regard to whether or not others endorse it. And as they shrink
from investing their own view of social justice with such a special, peremptory
status, of course, so they deny that status to the views of others. They agree
in each renouncing any form of crusading vanguardism or sectarianism in
favor of an attitude of mutual respect or forbearance.
This disposition of forbearance is unsurprising in view of an assumption,
shared by theorists on all sides, about the more or less egalitarian character
of justice and about the more or less egalitarian character of their rival the-
ories. It is axiomatic for all, and manifestly axiomatic for all, that a just basic
structure must treat citizens as equals in their relations with one another. If
the defenders of rival theories are forbearing in their attitudes towards one
another, that is because they assume that each of them is seeking an inter-
pretation of what such treatment requires (Dworkin 1978). Thus, there is
no suggestion that they would be forbearing in relation to a social theory—
it could scarcely be called a theory of justice—that dismissed out of hand,
or clearly downgraded, the claims of some.
The disposition of forbearance among theorists of social justice may be
grounded in any of a variety of considerations: for example, that anything
short of this disposition is a recipe for civil war; that no system of social justice
will be stable if it is imposed on people without regard for their attitudes; that
under any plausible view of social justice respect for the views of others is itself
an essential requirement; that the best scheme of social justice is likely in any
case to be a compromise between rival theories; and so on. I assume that for
one or a number of such reasons, every contemporary theory of justice presents
itself in the forum of public debate as a theory offered for the consideration of
others, not as a blueprint to be coercively implemented, regardless of people’s
attitudes towards it. And I assume that this is as it should be: that the consid-
erations rehearsed do indeed support such a disposition of forbearance.8
This disposition of forbearance is certain to be triggered in real-world
scenarios. For it is a commonplace that in almost every society there are
8
This is to favor the democratic line in resolving Richard Wollheim’s (1962) paradox
in the theory of democracy.
14 Philip Pettit
likely to be many different views on what social justice requires. Whatever

the society you belong to, there are bound to be some laws or some policies
that you yourself will take to be socially unjust (Nagel 1987). This is mani-
festly true in pluralist, multicultural societies today but it is likely to hold
in any organized polity. The circumstances of politics, as Jeremy Waldron
(1999) argues, are essentially circumstances of disagreement, in particular
disagreement on matters of social justice.
The commitment to mutual forbearance among theories of social justice
combines with the fact of widespread disagreement about which theory is
correct to generate a problem. It implies that from within the perspective of
any theory of social justice, there is an issue as to what should transpire in
any real-world society where people differ about the best theory. Each the-
ory may endorse the idea that there should be continuing, respectful debate
about the different elements on which different views divide. But in the
meantime, the society is going to have to put various laws and policies in
place, as a matter of practical necessity. So what is to be done over that
period—in effect, the indefinite future—according to competing but for-
bearing theories of social justice?
Those theories may not offer any explicit answer to this question but
their general commitment to mutual forbearance supports a specific com-
mitment, given just the possibility of disagreement, to an ideal of political
justice. Under this ideal, the society for which rival theories would make
different recommendations ought to establish a decision-making process for
identifying and implementing suitable compromises between them. In par-
ticular, it ought to establish a process that is not tilted in favor of any one
side in the debate. True to their shared assumption as theories of social
justice, they must all hold that any just political process ought to treat all
citizens—all adult, able-minded, more or less permanent residents—as equals;
or at least treat as equals those who are willing to live with one another on
the same terms, not taking themselves to be special in some way. The com-
promise regime that such a just process would establish might not answer
to the demands of any single theory of social justice, and might not resolve
the theoretical differences between theories. But it would resolve those dif-
ferences for practical purposes—it would accommodate the differences, as
I shall say—and would presumably remain open to development in light of
further theoretical convergence.9
9
John Rawls (2007, 2) might be taken to endorse the line presented here when he
writes: “Political philosophy can only mean the tradition of political philosophy; and in
a democracy this tradition is always the joint work of writers and of their readers. This
work is joint, since it is writers and their readers together who produce and cherish works
of political philosophy over time and it is always up to voters to decide whether to
The commitment to the idea that there ought to be a just process of deci-
sion-making for accommodating their differences is inescapable for theories
of social justice. If the forbearing defenders of different theories were to reject
the commitment, then they would have to embrace one of two unpalatable
alternatives. The first would be to reject any semblance of realism about
politics and argue that short of theoretical convergence, there should be no
practical accommodation among different views of social justice; the debate
should continue indefinitely. The second alternative would amount to
rejecting any suggestion of idealism in politics; it would accept that, no just
accommodation of differences being possible, each society is entitled to let
non-normative forces shape how it goes, indifferent to people’s rival views
of social justice.
To accept either of these views would be to take the theory of social jus-
tice to be irrelevant to public life. The first view would restrict normative
theorizing about social justice to the seminar room, which is the only loca-
tion that welcomes continuing, possibly endless debate. The second view
would despair of the capacity of normative theorizing to have any impact
on the life of a society. To endorse one or the other view would be to deny
any practical point to the theory of social justice, insulating normative
thinking about politics from the political world. Assuming that theories of
social justice can scarcely support this attitude—it would deprive them of
any application, after all—I conclude that they must commit to the idea
that each society ought to have a just process of decision-making to estab-
lish compromises between the rival views that emerge there.
The commitment ascribed here to theories of social justice is existential
in character, not substantive. It involves accepting, not that a particular
decision-making process should be established in the society for which
rival theories make recommendations, but that one or another just decision-
making process—maybe this, maybe that—ought to be put in place. For all
that we have said so far, theories of social justice may be silent on the form
that this process should take; on that matter they will presumably defer to
the theory of political justice. But as we shall see in the two following sec-
tions, there are other assumptions built into theories of social justice that do
put constraints on what form a theory of political justice ought to assume.
This line of argument suggests, then, that all forbearing theories of social
justice are committed to thinking that in circumstances of disagreement some
just process of decision-making—to be identified, presumably, in the theory
of political justice—ought to be introduced to accommodate their differences
in advance of whatever theoretical convergence may be expected. Each theory
embody their ideas in basic institutions.” Thanks to Ben McKean and Larry Udell for
drawing my attention to this passage.
16 Philip Pettit
is committed, as we might put it, to the idea that while it is theoretically

superior to alternatives, what is practically implemented in the society ought
to be decided by a just process of decision-making. Thus, the message in each
theory is that notwithstanding its theoretical merits, it ought to be practically
implemented only if it is itself selected by a just process of decision-making.
The recommendations that the theory makes are provisoed, we might say, on
their being implementable in a way that satisfies political justice.
This observation does not establish a straightforward priority for political
justice in relation to social justice. For as there are differences among theo-
ries of social justice, so there are differences among more elaborate theories
of political justice: that is, among theories that identify different, more or
less detailed processes of decision-making as just. One theory may recom-
mend a plebiscitarian system of popular voting on every issue, for example;
another, a mixed constitutional system of popular decision-making under
which power is divided out among separate bodies; another, a system of gov-
ernment that recruits expert elites to different domains of decision-making;
another, a mix of such processes, varying between domains of policymaking;
and so on. The divergence among such theories of political justice means,
assuming forbearance, that according to each theory, what is implemented
ought to be determined in a just compromise between alternatives and that
it itself ought to be implemented, therefore, only if it is selected by a just
process of decision-making. Thus, elaborate theories of political justice have
the same provisoed status as theories of social justice.
This need not make for a problem, however, and it need not undermine
the priority of political justice. The divergence among elaborate theories of
political justice is consistent with the belief that there is some more basic,
just process of decision-making that ought to be used to accommodate the
differences between them, as it accommodates the differences between the-
ories of social justice. And so they must share with theories of social justice
a commitment to the idea that in every society for which they make pre-
scriptions there ought to be a process of basic political justice in place to
accommodate differences between them. This process would have to be
unambiguously required by political justice or would have to belong to a
family of processes any one of which would satisfy political justice.
3. THIS PROCESS OUGHT TO REQUIRE A FORM

OF DEMOCRATIC APPROVAL
We have argued that elaborate theories of political justice, and all theories of
social justice, are committed under the assumption of mutual forbearance
to holding that some just process of decision-making between rival proposals
ought to be established in any society for which they make prescriptions.

But we have said nothing about the sort of process that they require other
than stipulating that, being just, it ought in some intuitive dimension to
satisfy the requirement of treating all citizens of the society as equals. In this
section and the next I argue that as a matter of fact the theories impose two
further constraints on the process: first, that it should attract a form of
democratic approval and, second, that it should introduce a form of demo-
cratic control. I shall concentrate for simplicity on the way in which theories
of social justice constrain the politically just process to which they are commit-
ted but, as occasionally noted, the points to be made apply also to elaborate
theories of political justice.
Every theory of social justice, as we have seen, recommends under con-
ditions of disagreement that despite its own theoretical superiority, what
is implemented ought to be determined by a just decision-making pro-
cess: a process for selecting a fair compromise among rival views. It is the
forbearance of rival theories that commits them to the need, under con-
ditions of disagreement, for such a just process. But a further feature of
theories of social justice implies that this just process should have a par-
ticular character: it should be a process that attracts what I describe as dem-
ocratic approval. The further feature is that every theory of social justice
makes recommendations rather than giving directives or commands and, in
particular, that it makes public recommendations: that is, recommenda-
tions addressed in public to those who are able and entitled, by reasonable
criteria, to take part in the political process. I take these parties to be the
adult, able-minded, more or less permanent residents who, by our earlier
assumption, constitute the citizenry.
When I make a recommendation to you, I do so on the presupposition
that you are a fit addressee of the recommendation. You are able to under-
stand what I am doing in making a recommendation; you are able to con-
sider the pros and cons of the recommendation, interrogating me insofar as
this is possible or appropriate; and depending on whether you endorse or
reject the recommendation, you are able to choose between acting on it or
not. Thus, in making a recommendation I ascribe a variety of capacities to
you by way of a presupposition. I do this moreover in awareness that that
presupposition is manifest to both of us; we are each linguistically compe-
tent enough to know what I presuppose, to know that we each know this,
and so on. Thus, in making any recommendation I communicate the con-
tent of my presupposition to you just as clearly as I would have done had
I spelled out the list of capacities I ascribe.
This observation means that when I recommend that you take a certain
course of action, X, I communicate two messages: first, the foregrounded,
semantic message that you should do X and, second, the backgrounded,
18 Philip Pettit
pragmatic message that you are someone who can understand, assess, and
act on the recommendation. Communicating both messages at once, my
speech act amounts to an invitation to you to do X on the basis of the rea-
sons presumptively supporting it; the suggestion is that you are able, enti-
tled, and even perhaps obliged to take that line.
The pragmatic message in a recommendation is not capable of being
cancelled. I can hardly say: I recommend that you do X but I do not mean
to suggest that you are capable of appreciating what a recommendation is.10
And so the presence of the pragmatic message puts constraints on the
semantic content that the recommendation can carry. I may recommend
that you should do X, where I can and do assume—although perhaps mis-
takenly—that X is within your deliberative control: something that you
can do or not do, depending on your assessment of the pros and cons. But
I cannot recommend responses that do not fit this constraint. I cannot rec-
ommend, for example, that you should blush or fall in love or be intimi-
dated, where it is manifest to all that those responses lie outside your
deliberative control.11
The fact that each theory of social justice makes a recommendation or set
of recommendations rather than issuing a series of directives or commands
carries a clear pragmatic message. It presupposes that an addressee of the
theory has the capacity to understand what is recommended, to assess the
recommendation on the basis of the pros and cons, and to try to act on it in
the event of a positive assessment. And it communicates that presupposi-
tion to addressees, implicitly inviting them to exercise the presupposed
capacities in relation to the recommended course of action. Whatever the
semantic content of the utterance—whatever the response recommended—
this pragmatic message is absolutely unambiguous.
But the recommendation or set of recommendations made by any theory
of social justice is quite distinctive in character, since it is offered in public
10
On the difference between pragmatic messages that can be cancelled and those that
cannot, see Jackson (1987).
11
On the assumption that it is manifest that you cannot deliberatively control blush-
ing or falling in love or being intimidated, such a recommendation would involve a
pragmatic inconsistency, as it is called: an inconsistency between the pragmatic and seman-
tic message of the utterance. The paradigm of such an inconsistency, due to G. E. Moore,
is “p but I don’t believe it.” This utterance communicates pragmatically that I hold by the
belief that p but communicates semantically that I do not believe that p. There is no
inconsistency in its being true that p and that I do not believe it, as there is no inconsist-
ency in my recommending that you X and its being impossible for you to X. But there is
an inconsistency in my implicitly communicating that I believe that p while claiming
explicitly that I do not believe it, as there is an inconsistency in my implicitly communi-
cating that you can deliberatively control X-ing when it is manifest to all that X-ing lies
beyond your control.
to all the citizens of any society that falls within its purview; this society may
or may not be the theorist’s own. If a theory were to avow the aim of offer-
ing advice only to an administrative elite, as Renaissance political tracts
offered advice only to princes, then by current standards it would under-
mine its own credentials and fail to count as a theory of justice proper
(Habermas 1971). Indeed it would even run into conflict with the medieval
principle according to which what affects all ought to be debated and
approved by all: Quod omnes tanget ab omnibus tratactari et approbari debet
(Congar 1958).
The publicity of the recommendation offered impacts on the pragmatic
message of the theory and means that it communicates a very specific form
of invitation to addressees. Each theory communicates the message that all
the citizens it addresses can appreciate and try to act on the sort of recom-
mendation that it puts before them. And because it is offered as a public
or manifest matter—that is, in such a way that each knows that each is
addressed, each knows that each knows this, and so on—it communicates
the message that they can appreciate and try to act on it jointly. Thus each
theory communicates the message that it is within the competence of the
public or citizenry as a whole to collectively assess and act on the recommen-
dations made: to establish the institutional means whereby those recom-
mendations can be considered and, if approved, implemented. It amounts,
in effect, to an invitation to the citizenry to get together and determine a
shared view of the proposal offered.
We saw earlier that each theory of social justice is committed to the idea
that under the conditions of disagreement that obtain in every society, there
ought to be a just decision-making process in place for establishing a compro-
mise between competing views. Once we recognize the pragmatic message
that each theory communicates in virtue of making public recommendations
to all the citizens of a relevant society, we can see that this commitment to
the need for a just decision-making process has a more determinate charac-
ter than registered earlier. It amounts to a commitment to the need for a just
decision-making process in which each citizen, each member of the public,
is invited to participate in a deliberative manner.
By this account, then, the commitment made in a theory of social jus-
tice is that each society ought not only to put in place a just process for
accommodating differences between competing theories but that it ought
to put in place a process that all the members of the society are ready to
accept. This requirement strengthens the earlier requirement that the pro-
cess for accommodating differences should be just. The requirement it puts
in place—that all be ready, presumably with good reason, to accept the
process—entails that it is just, as the earlier claim maintained, but not the
other way around.
20 Philip Pettit
The requirement introduced may seem to be excessively demanding,

however, suggesting as it does that citizens in every society should address
the issue of what process to establish in a single deliberative episode, should
consider the various candidates for the role of overall process, and should
unanimously support one or another candidate. But that suggestion is
misleading in two respects.
First, the requirement is not that there should be some single process that
all accept for every policy area. It need only be that for each area of social-
justice policy there should be some process accepted by all, where that
process—and even the personnel operating the process—may differ between
areas. The processing required may be distributed, in other words, not cen-
tralized.12 And second, the requirement is not that in any area people in
general, or the personnel they authorize, should actively consider the alter-
native processes available and opt for a single candidate. It will be sufficient
that some salient candidate emerges in that area and that everyone is pre-
pared, without any special pressure, to acquiesce in following it. The accept-
ance that agents give to the process, in other words, may be virtual rather
than active: it may consist in their failing to object to a particular proposal
that they are able to object to, not in their actively seeking and selecting the
proposal they positively prefer.
Suppose the society faces issues of social justice in three policy areas:
border protection, crime prevention, and travel routes. The requirement is
not necessarily that there should be one single process in place to resolve
every issue in all three areas. It will be enough that in each area—or on each
issue in each area—people go along with a particular process, perhaps with
different agents running the process in each case; in that sense their pro-
cessing of decisions is distributive rather than centralized. But equally the
requirement is not necessarily that in each area or on each issue people
should actively consider all available processes and converge on one. It will
be enough if a salient candidate emerges and commands virtual acceptance.
No one objects to the process, even when they are in a position to do so:
even when they can exercise a veto without any special cost to themselves.13
12
The distributive character of the system, as envisaged here, allows for distribution in
two dimensions: first, in the processes employed in different domains of decision-making
and, second, in the agents—the representatives of the people, as it is natural to say—who
operate those processes.
13
Any veto will deny the person exercising it access to a shared process—and pre-
sumptively a shared benefit—as it will deny this to others too. I think of that cost as
general rather than special. While it may induce someone to settle for less than what they
consider best, it is very different from the special cost associated, for example, with being
bullied not to cast a veto or ostracized as a result of doing so. I assume that if the shared
benefit of the arrangement is sufficient to get people to accept a process, then that process
A salient candidate may emerge with someone’s making a proposal to

follow a particular process—say, majority voting—and with no one’s object-
ing to that process. Or it may emerge with someone’s making a substantive
policy proposal—say, to have a wall built around the border, to have a mili-
tia guard against crime, to impose a tax for building roads—and with no
one’s objecting to that policy. In this case, the process endorsed is one under
which any policy is to be favored that someone proposes—someone special,
perhaps, or anyone in the group—and no one rejects.14
This argument establishes a strong connection between endorsing a the-
ory of social justice and being committed to something approximating
democratic procedure, albeit a procedure that may be distributed rather
than centralized, and may involve virtual rather than active acceptance. The
sort of connection made is close to that which Amartya Sen (2009, 326)
registers when he notices, on epistemic grounds, that if making recommen-
dations of social justice appeals to people in their capacity to deliberate or
reason, then it presupposes that decisions about justice ought to be formed
and implemented democratically. “If the demands of justice can only be
assessed with the help of public reasoning, and if public reasoning is consti-
tutively related to the idea of democracy, then there is an intimate connec-
tion between justice and democracy, with shared discursive features.”
But the connection to democracy that the argument supports falls short of
a commitment to full democratic control. All that is strictly required is that
the process involved should attract the approval or acquiescence of each, not
that it be a continuing process in which each maintains a presence and an
influence. Thus, for all that is strictly required by the argument, the mem-
bers of the society might opt in one or another area of policy, or even in all
areas, for irrevocably investing decision-making power in a continuing
dynasty or party, or in an agency that is selected at regular intervals on some
non-democratic basis: say, on the basis of virtue and expertise. They might
must promise to serve them reasonably well. I set aside the problems that strategic
game-playing might raise.
14
This virtual mode in which a group endorses a process is logically prior to any
active mode of endorsing it: say, the way members endorse a process when they vote
unanimously in favor. Even if members vote unanimously in favor of a process, this will
carry weight only insofar as they virtually endorse the unanimitarian process itself. And
if they seek to endorse that unanimitarian process in some independent voting procedure,
then any support they provide will carry weight only because of the virtual acceptance
of that procedure. The lesson is that there is no collective endorsement of any proposal,
including the proposal to follow a certain process, without the purely virtual endorse-
ment of some process of endorsement. It parallels the lesson that Lewis Carroll (1895)
taught us in the case of reasoning: that there is no way in which an individual can derive
conclusions from premises without the purely virtual endorsement of some principle or
rule of inference.
22 Philip Pettit
hold, rightly or wrongly, that such an arrangement promises a good prospect

for identifying compromises between different theories of social justice.
It may seem implausible that a theory of social justice should treat its
addressees as capable of jointly considering and acting on its recommenda-
tions, or on the compromise recommendations extracted from a bunch of
competing theories, yet accept that they are entitled to alienate this power
to an independent body. It was this sort of implausibility that led Rousseau
(1997, I.4) in writing The Social Contract to rail against the idea, championed
by earlier thinkers like Bodin, Grotius, and Hobbes, that a self-governing
people might irrevocably decide to give themselves over to the power of an
individual sovereign. “To say that a man gives himself gratuitously, is to
say what is absurd and inconceivable; such an act is null and illegitimate,
from the mere fact that he who does it is out of his mind. To say the same
of a whole people is to suppose a people of madmen; and madness creates
no right.”
Rousseauvian rhetoric aside, however, there is no strict reason why the
commitment embodied in theorizing about social justice—in supporting
public recommendations about how a society ought to organize itself—
should not allow the citizens addressed in the theory to opt for outsourcing
the capacity on a permanent basis. Hence, I think the most that the argu-
ment of this section can establish is a commitment on the part of theories
of social justice to the need for a once-for-always process of decision-making
that is approved by each, not to the need for an ongoing process in which
each retains a certain presence and influence.
The line argued here about theories of social justice applies also to elaborate
theories of political justice: theories that seek to lay down detailed processes
of decision-making, ranking different possible structures against each other.
Being forbearing theories, as we have seen, these resemble theories of social
justice in committing to the idea that there ought to be some basic, just
process of decision-making established in any society for accommodating
differences between rival theories. And being put forward as public theories,
manifestly addressed to all the citizens of any society on which they bear,
they must also resemble theories of social justice in presupposing that what-
ever process of decision-making is put in place, it ought to be one that
attracts the approval or acquiescence of each.
4. THIS PROCESS OUGHT TO INTRODUCE A FORM

OF DEMOCRATIC CONTROL
The forbearing character of theories of social justice—and elaborate theories

of political justice—argues for their commitment to the idea that in any
society there ought to be a just process for accommodating differences between

rival views. And the pragmatic character of those theories—their character as
theories that make recommendations rather than giving directives—argues
for their assuming that whatever process of decision-making is established, it
ought to attract democratic approval or acquiescence. But there is a third
feature shared by these theories and it supports an argument for the further
specification of the process of decision-making that is to accommodate dif-
ferences between them. This feature consists in the fact that they are all
theories of justice and that justice is a robustly or modally demanding good
(Pettit 2015).
The goods or values that are regularly invoked in normative theory vary
in the strength of the requirements they impose. I do not mean that they
make more or less exacting demands on our incentives and efforts, though
they certainly do vary in that way. I have in mind a variation in how far they
make demands, not just that things should be actually thus and so, but also
that they should be thus and so in counterfactual or modal scenarios: that
they should remain thus and so, for example, in scenarios where things
change in one or another respect.
Take a value like pleasure or success. To realize either of these goods in
your own life, all that is required is that actually you come to enjoy pleasure
or actually you manage to achieve success. Even if you realize that good
quite fortuitously—even if you would have failed in the event of things
being different in the slightest degree—that does not matter. Pleasure is
pleasure and success is success, no matter how contingent its realization;
indeed, it may be all the sweeter for being snatched in the luckiest of strokes
from the jaws of chance. And this is true, not just with achieving pleasure
or success in your own life, but also with achieving it on behalf of others
or on behalf of a group. If we can succeed in identifying a new antibiotic,
powerful against currently resistant strains of bacteria, none of us is going to
worry about the fact that it was sheer chance that led to the discovery.
But things are different with other values. Take the value of freedom,
even as conceived in the work of an orthodox thinker like Isaiah Berlin. He
argues that in order to enjoy freedom in a choice between certain options,
say X and Y, it must not only be the case that you are not interfered with in
taking the option you prefer, X; it must also be the case that had you pre-
ferred Y then you would not have suffered interference with that option
either (Berlin 1969; Pettit 2011). The argument is that if it is enough for
freedom of choice that you get what you prefer then, absurdly, you could
make yourself free in a choice where you are currently frustrated by adapt-
ing your preferences so that you come to want what you can get rather than
getting what you want. Confined to prison, and wanting to live in the out-
side world, you could get to be free by thinking about the benefits of prison
24 Philip Pettit
life—regular meals, a roof over your head, and so on—and getting your
preferences to shift appropriately.
Freedom, by this argument, is robustly demanding insofar as it requires
that if you are free in a choice between certain options like living inside or
outside prison then those options must each be open doors. You must be
able to choose according to your actual wishes and it must be the case that
had your wishes been otherwise you would also have been able to choose
according to those wishes; you must enjoy non-interference in either event.
But this conception of freedom is not the only one that makes it into a
robustly demanding good. A more traditional construal holds that freedom
is robustly demanding in another measure also.
On this traditional reading, freedom in a choice requires that your being
able to choose according to your wishes, regardless of what you wish, should
not be dependent on anyone else allowing you that license. If you were
dependent on the goodwill of another for being able to choose according to
your own will, then the other’s will would be in ultimate charge, not your
will; and in that sense you would be unfree (Pettit 2014).15 In order to enjoy
a properly free choice, the options must not only be open doors; there must
be no doorkeeper on whom you depend for leaving them open. You must
be able to choose as you wish, not only regardless of what you want to
choose, but also regardless of what others want you to choose or be able
to choose.
As it is with freedom, so it is with any form of justice. Take the private
justice, first of all, that you may experience in your dealings with me. You
do not enjoy justice at my hands if I merely happen, as luck would have it,
to satisfy the actual claims that you make on me; I must be disposed to
satisfy those claims robustly across variations in my particular inclinations.
I can act justly towards you in the absence of such a disposition—I can
satisfy your claim in justice—but without that disposition I do not count
as a just person in my interaction with you: I do not give you the benefit
of justice.
Now consider the justice, social or political, that you may hope to expe-
rience in your relations with your community as a whole. As in the other
case, you will not enjoy this sort of justice—this public as distinct from
private justice—if it is merely good luck that you are given satisfaction of
your claims, whatever they are taken to be. But neither will you enjoy this
sort of justice if that depends on others in your society being suitably virtu-
ous: that is, being disposed to grant you satisfaction of your public claims,
as they may be disposed to grant you satisfaction of your private. It must be
15
This construal is associated, in my view, with the long republican tradition. (See, for
example, Pettit 1997; Skinner 1998; Pettit 2012.)
that the institutions of the society ensure your claim-satisfaction actually

and across the range of possibilities where others happen to turn against you
or happen not to have the personal virtue of justice. Public or institutional
justice is doubly demanding in the way in which freedom, on traditional
construals, is doubly demanding.
The robustness of the requirements of public justice, social or political,
shows up in the idea, in contemporary parlance, that to enjoy justice is to
be able to claim just treatment as a matter of right. We as a community do
not grant or validate your claim to just treatment as a matter of right merely
because we happen not to violate that right in actual circumstances. We
must treat you as a rights-holder, being each constrained to treat you justly
in a way that does not depend on its answering to our independent inclina-
tion or even our personal virtue.
The idea that justice requires the robust satisfaction of relevant claims,
not only their fortuitous fulfillment, has a long history. It appears already as
a principle of justice in the Digest of Roman Law, produced under the
Emperor Justinian in the sixth century bce: Justitia est voluntas constans et
perpetua jus suum cuique tribuendi; “Justice is the steady and enduring will
to render unto everyone his right” (Watson 1985, I.1.10). This requirement
is defended in those very terms by Thomas Aquinas (1958, II-II, Q58, 1) in
the thirteenth century. And a variant is upheld by Thomas Hobbes’s (1994,
Ch. 15) in the seventeenth, when he comments that a just man is “he that
taketh all the care he can that his actions may all be just, an unjust man is
he that neglecteth it.” These observations all bear on private justice, of
course, but they support the line taken on public justice. A constant and
perpetual will on the part of the community to grant your claims would
surely require the satisfaction of those claims to be independent of whether
or not others are possessed of personal virtue.
The idea that justice is a robustly demanding good has implications for
how a system of social justice ought to be legally structured. Your claims in
social justice must be robustly satisfied over variations in the dispositions of
those who are to satisfy those claims; you would not enjoy justice proper if
the satisfaction of your claims turned on whether they happened to feel like
satisfying them or happened to be virtuous. If the society is to be structured
so that you enjoy social justice, the law must not leave the satisfaction of
relevant claims up to the benevolence of others; it must constrain others so
that they have little or no choice but to satisfy those claims.
Some of your claims in social justice will be held against other individ-
uals, as in claims against violence, deception, manipulation, harassment,
free-riding, and the like, and in these cases the law had better not leave how
they treat you up to the discretion of your debtors. Others of your claims
in social justice will be held against government officials, as in claims to
26 Philip Pettit
social protection or insurance in the event of suffering penury, unemploy-

ment, ill health, or a variety of such misfortunes. And in these claims too
the law should not leave you to the mercy of the counter-clerk or any
individual official; as far as possible it should establish rights for you in
these areas.
But not only must the law that delivers social justice constrain the indi-
viduals or agencies that deal with you directly on these fronts. It must also
constrain the lawgivers and policymakers—the officials of the state—in
how they choose to frame or apply that law. Suppose that the law were in
the hands of a single individual, say a benevolent despot, or subject to the
say-so of a particular elite. In that case, whether or not the law protected
you or your group—whether or not it constrained how you were treated in
general or in particular instances—would still depend on the will of the
dictatorial individual or elite, in particular their attitude towards you and
your group. And so in this case too you might be lucky enough to be treated
justly but you could not be said to enjoy social justice.16
The lesson here is illustrated by the position of Jews in Nazi Germany
prior to the imposition of anti-Semitic measures but at the point where it
was clear that the government did not see them as equals with others and
was just waiting for an opportunity to deprive them of their rights without
excessive diplomatic or judicial risk. In that brief period even Jewish citizens
who were treated according to the law by other citizens and by state officials
did not have the tenure on their rights that social justice strictly requires.
They could not say: this is a more or less just society and we enjoy social
justice within it; we enjoy access to relevant benefits on the basis of rights
that we hold as firmly as anyone else in the society.
Let us agree on the basis of these considerations that social justice requires
the robust satisfaction of relevant claims, whatever they are taken to be; it
requires having robustly demanding rights to just treatment. This means
that when defenders put forward rival theories of social justice they presup-
pose, or ought to presuppose, that the approved process of decision-making
for adjudicating rival claims should give a robustly demanding status to the
16
This observation means that I dissent from the position G. A. Cohen (2008, 269)
gestures at when he sets up a contrast between the “fundamental principles of justice”
with which the philosopher ought to be concerned and the “rules of social regulation,”
which bear on the presumably empirical matter of how they are best implemented in
actual circumstances. Justice requires rules of social regulation as an inherent matter, on
the account given here, not just as instruments for realizing prescribed patterns. It is
those rules or institutions that are going to determine whether or not the patterns that
our principles of justice identify are robustly in place. And unless they are robustly in
place, the realization of those patterns does not provide justice. In this criticism of
Cohen’s view, I join with Charles Larmore (2012), and Jeremy Waldron (2013), though
for somewhat different reasons.
compromise claims it supports. The process should be sufficient to ensure

that the citizens of the society can rely on their claims being satisfied across
variations both in the dispositions of those who deal with them directly and
in the dispositions of those who frame and apply the law.
What might make a process sufficient to ensure this? There is no abstractly
compelling criterion that is likely to provide an answer that holds across con-
tingencies of culture, history, and technology. I shall assume here that a process
will be sufficient to ensure the required robustness of claim-satisfaction to the
extent that it passes “the eyeball test”: by the most demanding of local stand-
ards it enables people, absent timidity, to look others in the eye without reason
for fear or deference (Pettit 2012; 2014). It should be clear that Jews in even the
early days of Nazi government were not in a position where they could pass this
test. The law may have continued to deal with them as equals but it had already
ceased to provide them with the security that justice demands.
What exactly do theories of social justice mandate, in requiring that the
process for accommodating differences between rival theories should pro-
vide for the satisfaction of everyone’s assigned claims with a suitable degree
of robustness? That the process has been approved by all, as we argued in the
second section that it must be, will not guarantee that it meets this require-
ment. For in allowing that a particular dynasty or party or elite might be put
irrevocably in charge of framing or applying the demands of public justice,
this act of approval would enable a particular individual or group to act for
its own advantage, and to deprive some others of their rights in justice. And
were the group in power able to do this, then no matter how benevolent it
proved to be, it would not deliver the satisfaction of relevant claims with the
robustness required for social justice.
The question of what the process of accommodating different views of
justice requires in institutional terms is not one that it is possible to discuss
here at appropriate depth or in appropriate detail. But it is hard to see how
any process that departed from a basic democratic specification could do
the job satisfactorily. The process would have to ensure that absent a right
of exit, the citizens of any coercive regime ought each to be able to avail
themselves of the only effective alternative for self-protection. They ought
to be able to exercise their voice in a manner sufficient to guard against the
possibility that any one individual or any one subgroup is not treated as
equal with others within the local system of social justice. But a process that
enabled citizens to exercise their voices in this way, giving them each an
equal place in determining or controlling the laws and policies established
in the society, would give kratos or power to the demos or people, and give it
with due equality. And it would be hard to deny the title of a democracy to
any regime that satisfied this constraint in significant measure, whatever the
institutional means chosen for meeting it.
28 Philip Pettit
The only basis for questioning the need for democracy in this sense is the
thought that the process for accommodating differences between theories of
social justice might be taken out of human hands altogether, or at least out
of the hands of intentional designers. As a matter of logical possibility, an
impersonal or non-intentional process might establish a compromise scheme
of social justice that gave each the same claims against others and gave them
on a basis that was robust over shifts in individual inclination or bias. But
there is no real-world possibility or prospect of such a robotic process,
although variations on the idea have beguiled a number of thinkers. Perhaps
it was a fascination with this idea, for example, that explains why Frederic
Hayek (1988) looked for a system of decentralized, common law courts—
surely a fantasy on stilts—that would generate a bare form of social justice
as by an invisible, unintentional hand.
This line of argument leads me to support the third of my progressively
stronger theses. Theorists of social justice are not only committed to the
idea that there ought to be a process for accommodating differences between
rival theories that is just and has democratic approval. As advocates of jus-
tice, robustly understood, they are also committed to the claim that the
process ought to embody a degree of democratic control, giving people an
equal voice in exercising control over how that process goes. This constraint
is stronger again than the previous two. A process that gave people some-
thing like an equal voice, satisfying this requirement, would have to satisfy
the earlier requirements too but not the other way around. It would pre-
sumably attract the approval of all—or at least all who are willing to live on
equal terms with others—and, treating all as equals in that respect, it would
count moreover as just.
As with the argument for the first and second constraints, this argument
about the commitment of theories of social justice applies also to elaborate
theories of political justice. They too are committed, as we have seen, to the
idea that there ought to be a just process for accommodating differences
between rival theories and, in particular, a process that attracts everyone’s
approval. But because they are theories of justice, albeit political rather than
social justice, the decision-making process that they require has to involve a
mode of framing and imposing law that gives each an equal voice and
thereby ensures that the satisfaction of people’s claims of political justice is
suitably robust.
The reference to theories of political rather than social justice prompts a
final observation that bears on the sort of democratic control that theorists
of robust social justice must require. If the political process were subject to
the say-so of a particular individual or dynasty or elite, as we noticed, then
it could certainly not deliver social justice with the required robustness.
Such a failure of political robustness would also affect social robustness,
even if the governing agency happened to grant people’s claims in social

justice; it would mean that the satisfaction of those claims was not really
robust after all (Pettit 2001). It is worth noting, in conclusion, that a similar
lesson bears on the way in which democratic control is structured.
Under a democratic constitution, whatever form it takes, the people as a
whole assume the profile of a corporate agency (List and Pettit 2011;
Pettit 2012, Ch 5). It is ultimately the corporate people that acts when
action is taken under that constitution by one or another representative
body or individual, or indeed by individuals acting jointly at the polls. But
now suppose that the constitution allowed the electorate to change the con-
stitution itself by a majority vote. That would mean in abstract principle
that the electorate could deny equal status, say equal voting rights, to some
of the members. And it would mean in general practice that many members
of the community—say, a salient, fixed minority—would not be able to
look others in the eye without reason for fear or deference. The electorate
would relate to individuals as an omnipotent force in the way in which
Rousseau allows the sovereign assembly to relate to individual citizens.
It should be clear that a democratic constitution that gave the electorate
such power over individuals might fail the eyeball test and not satisfy the
robustness requirement on just political process. And so the final implication
of our considerations in this fourth section is that not only must theories of
social justice, and indeed elaborate theories of political justice, commit to
the idea that there ought to be a democratic decision-making process for
accommodating differences between those theories. They ought to commit
us also to the idea that this process should restrict the power of any single
agent, even the electorate as a whole, when that agent acts in the name of
the corporate people.17
With this final lesson spelled out, it ought to be clear that the divide
between theories of basic and elaborate political justice, with which I have
been working in this chapter, may not be as wide as suggested. As we begin
to think about the basic process that ought to be in place, recognizing the
robustness that it must display, the elaborate theories of political justice
that remain in place as rival and plausible candidates are likely to shrink
in number. The basic design specification is very demanding, after all: the
process must provide each with an equality of voice sufficient, by the eye-
ball test, to make the delivery of political justice suitably robust; and so it
must be entrenched against the power of any agent or agency, even the
17
For an insightful discussion of democracy as a robustly demanding value, see
Southwood (2014). Where I invoke the eyeball test here for testing both social and polit-
ical justice, I invoke the related tough-luck test for checking political justice in Pettit
(2012, 2014), restricting the eyeball test to social justice.
30 Philip Pettit
electorate, that acts in the name of the people. There are many different
ways in which that specification might be institutionally implemented, of
course, and the process has to be able to accommodate differences between
these elaborated proposals. But the specification is still sufficiently demand-
ing to put many theories of political justice off the shortlist of plausible
candidates. It holds out the prospect that only a small family of models
may be available to exemplify what a basic process of political justice might
look like.
5. THEORETICAL AND PRACTICAL IMPLICATIONS
There are two theoretical implications of the argument in this chapter,

each underlining a way in which the theory of political justice—in par-
ticular, the theory of democracy—has priority over the theory of social
justice. The argument means, first, that the theory of political justice has
a certain formal priority and, second, that it has a substantive or material
priority as well. The formal priority shows up in the fact that the basic
theory of political justice—the theory bearing on the process of accom-
modation that ought by all accounts to be in place—is not provisoed in
the way in which theories of social justice, and elaborate theories of polit-
ical justice, are provisoed. The material priority shows up in the fact that
what that theory of political justice requires is likely to involve the satis-
faction of conditions often associated with quite demanding theories of
social justice.
Turning to the formal priority, we know from previous discussions that
any forbearing theory of social justice—and any forbearing theory of elab-
orate political justice—recommends that it be implemented only under
the proviso that it is selected by a just, democratically approved, and dem-
ocratically structured process of decision-making. Put in more everyday
terms, this means that when we argue as philosophers about what justice
requires in these areas, our position vis-à-vis our audience is that of an advi-
sor of citizens. Whether or not we actually belong to the society, we adopt
the perspective of insiders and argue inter cives, among citizens, that we the
people as a whole, or the government that acts in our name, ought to do this
or that. And as we make recommendations in that spirit, we recognize that
whether our proposals are to be accepted is up for general determination,
not something on which we in particular can issue dictates. We do not
claim any privileged, philosophical place in debating with our fellow citi-
zens (Walzer 1981). We expect to command their assent only to the extent
that we can reach and move them with our arguments (Marti and Pettit
2010, Ch. 5).
But whereas we put forward proposals in social justice—and in the elab-

orate reaches of political justice—under a democratic proviso of the kind
described, we assume a rather different position in relation to other citizens
when we pronounce as philosophers or theorists on matters of basic politi-
cal justice: in effect, on matters of basic democratic process. Suppose we
hold, to rehearse some themes canvassed earlier, that on our view of basic
political justice a society ought to have a decision-making process that
meets the following conditions: it is just in treating all as equals; it attracts
the approval or acquiescence of everyone; and it gives each an equal voice in
determining what the society does. Do we make that proposal under the
proviso that it itself is democratically endorsed? I argue not.
Suppose that operating with the sort of process recommended the peo-
ple reject its use in future cases—however unlikely that may be, given that
it is supposed to be suitably robust. Suppose, for example, that it is settled
under the process without objection from anyone—even, we may sup-
pose, without objection from women—that henceforth only adult males
should have the vote. Do we say that that is fine and that we have no right
as philosophers to question the wisdom of the people? Do we say that
democracy should be allowed to define democracy, as we allow that it may
define various demands of social justice? Surely not. At this level of nor-
mative theory, we stand our ground, arguing that a concern for people’s
rights to decide their fate democratically—a concern in that sense for their
collective autonomy—makes it impossible to endorse their democratic,
presumptively autonomous decision to set aside democratic rule. Here,
the philosophical spade hits bedrock, in Wittgenstein’s image. Here, there
is solid ground on which political philosophy can build without deference
to democratic will.
These considerations should help to explain the sense in which the the-
ory of basic political justice—in effect, the theory of democracy—enjoys a
formal priority over the theory of social justice, and indeed over the elabo-
rate theory of political justice. But it is worth noticing also that the theory
of democracy, as that has been foreshadowed here, promises to have a cer-
tain material priority as well. On the account sketched, it requires people
to have equal voices in a process for determining law and policy, and in
particular equal voices within a process that robustly delivers satisfaction
of the demands of social justice that it licenses. But if a society is to meet
this requirement, then it is likely to satisfy a variety of constraints bearing
on social justice.
The requirement is unlikely to be satisfied, for example, unless conditions
like the following are met. The citizens of the society are sufficiently edu-
cated and informed to be able to play a role in democratic decision-making.
Public life is sufficiently egalitarian to underwrite the claim that people have
32 Philip Pettit
equal voices: they are equally positioned to be heard—perhaps via repre-

sentatives—even if they cannot all be equal in the influence they happen to
exercise. The discretion of public representatives, assuming there has to be
representation, is constrained enough to ensure that they cannot put their
own interests before the interests of the people. The power of financial and
media elites is restricted to the point where they cannot exercise covert
influence for their own special interests. And so on.
As we envisage a society where conditions like these are realized then it
should be clear that it will have to exemplify a good deal, intuitively, of
social justice. The conditions are unlikely to be capable of realization in a
society where a decent infrastructure of law or education or commerce is
lacking, for example; or where people are not insured against social, medi-
cal, and legal misfortunes that can put them in the power of others; or
where those in relationships of asymmetrical dependence in the home or
workplace or public square are not protected against subordination and
subjection; or, of course, where there is only inadequate protection against
the harms—typically, the criminalized harms—to which individuals or
groups can expose innocent parties.
These observations suggest that the conditions that basic political justice
requires to be fulfilled are themselves sufficient to ensure that the society
meets many prominent standards of social justice (Habermas 1995; Forst
2002). It will have to have a well-functioning system of education, law, and
commerce. It will have to provide for at least a basic form of social insur-
ance against various maladies. It will have to support the vulnerable in
relationships of dependence. And it will have to establish a decent system
of criminal justice.
These comments bear on the theoretical implications of the argument
developed in earlier sections, underlining the dual sense in which political
justice enjoys priority over social. But what are the practical implications?
What is the upshot for how we should pursue political initiatives and policy
recommendations?
The priority of political justice that we mark in theory may seem to
argue in practice for refraining from advocacy in the realm of social jus-
tice until we have sorted out issues of political justice. But that would be
a serious mistake. Politics is mostly driven by divisions on ground-level
issues of social justice, as different individuals and groups within the
society rail, for example, at the unfairness of opportunities or outcomes
in this or that domain of life. The higher-level issues of political justice,
bearing as they do on how decisions ought to be reached on ground-level
matters, are inevitably more distant from people’s motivating concerns
and not as readily engaging. Thus, to try to prioritize political justice in
the practice of politics would be a hopeless and even worthless enterprise;
it would amount to trying to privilege the philosophical seminar over the

public forum.
But the priority of political justice does still teach a number of other
important, more or less practical lessons. It means that in putting forward
our proposals in social justice—as in most of the interventions we are likely
to take as activists within our own society—we should be alert to the pro-
cess whereby they are heard and judged in relation to alternatives and not
assume that that process will serve us well. It implies that for that or other
reasons we should address issues of political justice in their own right, and
not just focus on the issues of social justice that are more likely to command
popular feeling and to prompt popular mobilization. And it entails, finally,
that a good basis for arguing in support of various social policies may be,
not just that they are required as a matter of social justice, but that they are
even required to ensure that the political process lives up to its democratic
aspirations.
All of that said, of course, the philosophical seminar is probably the pri-
mary locus of operation for most of the readers of this chapter. And in our
philosophical work on issues of politics, itself of potentially practical inter-
est, the chapter does argue for redirecting some of our attention away from
the questions about social justice—for example, questions about equality
and welfare, domestic and international—that have dominated the disci-
pline in the last half-century or more. There is every reason why we should
give equal or even more attention to questions of constitutional and demo-
cratic design. To do this would not be to take a dramatically novel turn, of
course. It would be to return to the main topics of political argument in the
long history of our discipline.
Bibliography
Aquinas, T. (1958). Summa Theologica. Madrid, Biblioteca de Autores Cristianos.
Beitz, C. (1989). Political Equality: An Essay in Democratic Theory. Princeton, NJ,
Princeton University Press.
Berlin, I. (1969). Four Essays on Liberty. Oxford, Oxford University Press.
Carroll, L. (1895). “What the Tortoise said to Achilles.” Mind 4: 278–80.
Christiano, T. (2008). The Constitution of Equality: Democratic Authority and its
Limits. Oxford, Oxford University Press.
Cohen, G. A. (1989). “On the Currency of Egalitarian Justice.” Ethics 99: 906–44.
Cohen, G. A. (2008). Rescuing Justice and Equality. Cambridge, MA, Harvard
University Press.
Congar, M.-J. (1958). “Quod Omnes Tangit ab Omnibus Tractari et Approbari
Debet.” Revue Historique de Droit Francais et Etranger 36: 210–59.
Dworkin, R. (1978). Taking Rights Seriously. London, Duckworth.
34 Philip Pettit
Forst, R. (2002). Contexts of Justice: Political Philosophy beyond Liberalism and

Communitarianism. Berkeley, University of California Press.
Habermas, J. (1971). Knowledge and Human Interests. Boston, MA, Beacon Press.
Habermas, J. (1995). Between Facts and Norms: Contributions to a Discourse Theory of
Law and Democracy. Cambridge, MA, MIT Press.
Hayek, F. A. (1988). The Fatal Conceit: The Errors of Socialism. Chicago, University
of Chicago Press.
Hobbes, T. (1994). Leviathan. E. Curley (ed). Indianapolis, Hackett.
Jackson, F. (1987). Conditionals. Oxford, Blackwell.
Larmore, C. (2013). “What is Political Philosophy?” 10/3 Journal of Moral Philosophy:
276–306.
List, C. and P. Pettit (2011). Group Agency: The Possibility, Design and Status of
Corporate Agents. Oxford, Oxford University Press.
Marti, J. L. and P. Pettit (2010). A Political Philosophy in Public Life: Civic
Republicanism in Zapatero’s Spain. Princeton, Princeton University Press.
Nagel, T. (1987). “Moral Conflict and Political Legitimacy.” Philosophy and Public
Affairs 16: 215–40.
Pettit, P. (1997). Republicanism: A Theory of Freedom and Government. Oxford,
Oxford University Press.
Pettit, P. (2001). “Capability and Freedom: A Defence of Sen.” Economics and
Philosophy 17: 1–20.
Pettit, P. (2011). “The Instability of Freedom as Non-Interference: The Case of Isaiah
Berlin.” Ethics 121: 693–716.
Pettit, P. (2012). On the People’s Terms: A Republican Theory and Model of Democracy.
Cambridge, Cambridge University Press.
Pettit, P. (2014). Just Freedom: A Moral Compass for a Complex World. New York,
W.W. Norton and Co.
Pettit, P. (2015). The Robust Demands of the Good: Ethics with Attachment, Virtue and
Respect. Oxford, Oxford University Press.
Rawls, J. (1971). A Theory of Justice. Oxford, Oxford University Press.
Rawls, J. (1995). “Political Liberalism: Reply to Habermas.” The Journal of Philosophy
92: 132–80.
Rawls, J. (2007). Lectures on the History of Political Philosophy. Cambridge, MA,
Harvard University Press.
Rousseau, J. J. (1997). Rousseau: ‘The Social Contract’ and Other Later Political
Writings, trnsl Victor Gourevitch. Cambridge, Cambridge University Press.
Sen, A. (2009). The Idea of Justice. Cambridge, MA, Harvard University Press.
Simmons, A. J. (1999). “Justification and Legitimacy.” Ethics 109: 739–71.
Skinner, Q. (1998). Liberty before Liberalism. Cambridge, Cambridge University
Press.
Southwood, N. (2014). “Democracy as a Modally Demanding Value.” Nous 48.
Valentini, L. (2013). “Justice, Disagreement and Democracy.” British Journal of
Political Science 43: 177–99.
Waldron, J. (1999). Law and Disagreement. Oxford, Oxford University Press.
Waldron, J. (2013). “Political Political Theory: An Inaugural Lecture.” Journal of
Political Philosophy 21: 1–23.
Wall, S. (2006). “Rawls and the Status of Political Liberty.” Pacific Philosophical
Quarterly 87: 245–70.
Walzer, M. (1981). “Philosophy and Democracy.” Political Theory 9: 379–99.
Watson, A. (1985). The Digest of Justinian, Four Volumes. Philadelphia, University of
Pennsylvania Press.
Wollheim, R. (1962). A Paradox in the Theory of Democracy. Philosophy, Politics and
Society; Second Series. P. Laslett and W. G. Runciman. Cambridge, Cambridge
University Press: 71–87.
2
Voting and Causal Responsibility
Geoffrey Brennan and Geoffrey Sayre-McCord
1. INTRODUCTION
In the standard Rational Choice Theory (RCV) account of voting behavior,

the idea of a vote being “pivotal” plays a central role. Votes are pivotal when an
electoral option wins by one vote. In “pivotal” cases, the voters who vote for
the winning option (J) are both jointly sufficient and individually necessary
for J’s victory. Each voter composing that set can, on this basis, be described
as having “brought about” the victory of J. Each is causally efficacious in the
sense that if she had voted for a different candidate (and everything else had
remained the same), J would not have won. In all cases where the outcome
is non-pivotal, any individual voter, had she voted differently, would not
have altered the electoral outcome. Her vote would have been consequen-
tially irrelevant (in this sense) in all but the non-pivotal case.1 For this rea-
son, the ex ante probability of being pivotal is taken by RCV to be a crucial
parameter in each individual voter’s deliberations. It is this ex ante probabil-
ity (denoted by h henceforth) that influences:
how much information about alternative candidates it will be rational
for the voter to acquire;
1
There is a second-order complication that we should dispose of at the outset: J might
win the election following an exact tie among all voters. The decision over who wins in the
event of a tie might be determined by tossing a coin; and J, rather than K, might win that
coin toss. Each voter, in these cases, is such that had she abstained, or voted differently, she
would have brought about the victory of K. And in the case where J wins by exactly one
vote, if a voter had abstained, she might have brought about the victory of K via a tie-break-
ing mechanism. So, if voter A abstains rather than votes for J, in the first case, she would
cause K to win by a majority of 1; or, in the second case, she would give rise to a situation
in which K might win via a tie-breaking procedure. What this means is that the idea of
being pivotal is slightly different depending on whether the option under consideration is
voting for an alternative candidate or abstaining, and on whether the number of voters is
odd or even. Here, we shall treat the idea of being pivotal capaciously—allowing it to cover
all cases in which A’s voting behavior could influence which candidate wins.
Voting and Causal Responsibility 37
the extent to which decisions about which candidate to vote for will
be influenced by the voter’s individual self-interest;
and whether it will be rational to vote at all, given the stakes involved
in the election for the individual as she perceives them, and the
expected closeness of the race and the alternative actions to voting
available.2
In most large elections, clearly, h (even capaciously defined in the sense of
footnote 1) will be miniscule. Given this fact,3 rationality considerations (as
RCV understands them) can explain:
1. why individual voters do not expend much effort in acquiring rele-
vant information about candidates and their policies. After all, their
votes are very unlikely to make a difference to the outcome. (This is
the basis of Downs’ (1957) claims about electoral “rational ignorance.”
For a more recent treatment, see Caplan (2007).)
2. why considerations of individual interest as opposed to public
interest are likely to play a smaller role in decisions about how to vote
than in those same agents’ market decisions. After all, since the chance
that one’s voting decision will have an impact on the outcome is far
lower than in market contexts, where one’s choice is usually decisive,
the chance that it will have an impact on one’s individual interest is
correspondingly small. (This is the central upshot of the “expressive”
account of voting behavior—as analyzed, for example, by Brennan
and Lomasky (1993).)
3. why certain factors, and not others, are likely to influence aggregate
levels of voter turnout.4 After all, if the chance that one’s vote will
make a difference to the outcome is exceedingly small, the incentive
provided by the thought that one’s vote will make a difference will be
small as well.
2
It is perhaps worth emphasizing that the question of whether it is rational to vote or
not involves a deployment of the logic of rationality that is somewhat at odds with its use
in “rational choice theory.” Charges of irrationality involve argument from presumed
desires and beliefs on the one hand to actions that are inconsistent with those desires and
beliefs on the other. The rational choice theory approach involves argument from
observed behavior “backwards” to the beliefs and desires that are thought to motivate
that behavior. Here, rationality works as an analytic assumption rather than an issue to
be determined in any specific case. So, the presumption in RCV is (or ought to be) that
both voters and abstainers are rational: the question then becomes—what differences in
beliefs and desires of the two sets of agents would explain these behavioral differences?
The notion that one of these groups might be ‘rational’ while the other group is not is
alien from the spirit of the RCV enterprise—a fact which did not prevent some early
RCV theorists from speculating as to whether voting is “rational” or not. See for example,
Tullock (1967).
3
For example, in US Presidential elections, at most of the order of 1/12500 and
almost certainly rather smaller than this. See Brennan and Lomasky (1993) chapter 4.
4
The classic treatment is Riker and Ordeshook (1968).
38 Geoffrey Brennan and Geoffrey Sayre-McCord
An implication of RCV is that any voter who votes in standard large-

number democratic settings, thinking that her vote is likely to be causally
efficacious in determining the electoral outcome, is making a mistake.
Causal efficacy is possible: but its probability is so low that a rational voter
will not include that prospect as being a major, let alone a predominant,
“reason for voting.”5
In one sense, such conclusions could be seen as showing the power of
RCV logic. Means-ends thinking can indicate settings in which certain
things are unlikely to work as a means to an end because the connection
between the end (in this case the electoral outcome) and the means (the
individual’s vote) is too tenuous. Where there is reason to think one’s vote
won’t serve as an effective means to securing a particular electoral outcome,
the quest would then be for ends that could be appropriately cast as “rational
reasons” for voting. And, predictably, people have suggested ends other
than the electoral outcome that might provide people with reasons to vote
(e.g., a concern to do their duty, or to avoid substantial regret, or to serve
as an example to others). In each case, of course, the question will arise as
to whether voting is actually an effective means to the suggested end, but
assuming it is, one will have found reasons to vote that do not depend on
thinking that voting is an effective means to bringing about one’s preferred
outcome.
The RCV’s position, which is now standard in both economics and polit-
ical science, hinges on h (the ex ante probability that one’s vote will be piv-
otal) being the appropriate parameter for assessing an individual’s causal
efficacy when it comes to voting.
Alvin Goldman (1999), and more recently, with some variation,6 Richard
Tuck (2008), have challenged this idea. They grant that it matters whether
one’s vote might be the cause of victory. Yet, they maintain that the account
of causal efficacy accepted by RCV is wrong. On the RCV view, one’s vote
counts as causing the outcome if, but only if, one’s vote is pivotal—only if,
one were to vote differently, the outcome would be different. Against this,
Goldman and Tuck argue that what matters in thinking about whether
one’s vote might be the cause of victory is not whether it is necessary for the
outcome, but whether it is a part of what is, in the appropriate way, suffi-
5
Of course, expected benefit might be a reason for voting if the stakes for the voter
are huge—but then it is the size of the stakes rather than the probability of making a
difference to the outcome that is the primary “reason” to vote.
6
The variations are far from minor, but they share enough, we think, for it to make
sense to consider them together, not least their rejection of the idea that what matters is
that one’s vote will be pivotal.
cient for it.7 To take an example from Tuck, a police officer will have killed
a robber—will be the cause of the robber’s death—even if some other officer
would have shot the robber had the first missed. So the first’s accurate
shooting was not necessary in order for the robber to die. And a lifeguard
saves your live—is the cause of you not drowning—even if, had she not
saved you, another lifeguard standing by would have: so the first lifeguard’s
action was the cause of your being saved even though her saving you was
not necessary for you to survive. Similarly, on their view, one’s vote can be
a cause of an electoral outcome even in cases where there are more than
enough votes for victory, so even when one’s voting is not necessary for the
outcome (although Goldman and Tuck differ concerning just when one’s
vote is properly seen as a cause in such cases). As a result, they argue, h is not
the probability that deserves attention in thinking about whether one’s vote
will be causally efficacious; rather, what is important is the ex ante probabil-
ity that one’s vote will be among those that are, in the appropriate way,
sufficient for electoral victory. (In what follows, we refer to this as the
“Goldman/Tuck account.”)
So, while RCV and Goldman/Tuck agree that in thinking about whether
to vote a central question is whether in voting one will bring about, or
cause, the desired outcome, they differ concerning their understanding of
what it takes to be a cause. On the RCV view, the relevant test is counter-
factual: would the outcome have been the same even if you had voted dif-
ferently. If yes, then your vote is not pivotal, and so not (on this view)
causally efficacious. On the Goldman/Tuck account, even if your vote is not
pivotal, it will be causally efficacious if it is among the votes that are, in the
appropriate way, sufficient for the outcome.
There are some important differences between Goldman and Tuck con-
cerning just which sufficient set of votes matters in the context of voting. In
particular, on Tuck’s view, it is whichever minimally sufficient set actually
settled the election. The idea shows up nicely, as Tuck emphasizes, in a roll-
call vote, where at a certain point victory is secured by someone’s vote,
together with those before, even if those following in the roll call would
have voted in the same way.8 The same idea shows up, although with com-
plications due to the electoral college, in thinking about the causal efficacy
of votes cast on the West Coast or in Hawaii in US presidential elections.
Often, the outcome is settled, by the votes cast earlier, farther East, in a way
7
We write of being sufficient “in the appropriate way” because Goldman and Tuck
differ when it comes to identifying which votes are a part of what is sufficient for the
outcome. We highlight the difference below.
8
Tuck, p. 51.
that makes the later votes immaterial to the actual outcome.9 For other
votes as well, Tuck maintains, there is some subset of the votes cast for the
victor—a minimally sufficient set—that will have been causally efficacious,
because actually sufficient for the outcome.10 On Tuck’s reckoning, the ex
ante probability that one’s vote for a victorious candidate will be causally
efficacious—the probability that it will be in the minimally sufficient set—
is normally much higher than h, even in large landslide elections. For
instance, in an election with a million voters, where the victor is expected to
win with 60% of the vote, the probability that one will be in the minimally
sufficient set, g, is 0.66 (whereas h is vanishingly small). And this is true
even though, as is often the case, we cannot know which subset of the votes
actually proved to be the sufficient set.
Suppose in a two candidate race, J emerges as victor, receiving N votes
out of the total voting population of M. Now, since (M – N) voters voted
for the alternative (K), the minimal set of voters J required for victory was
in fact (M – N + 1). We do not know which particular voters compose that
minimal set. But Tuck supposes, for this example, that the votes were
counted in a particular order and there is indeed a determinate set of voters
who compose that minimal set. So there is a fact of the matter about which
voters these were. A, who voted for J, can reasonably ask what the probabil-
ity is that she (A) herself was a member of that minimally sufficient set.
And that probability is just the number of voters required for the mini-
mal majority divided by the actual number of J-voters—or (M – N + 1)/N.
Denote this probability, g. Tuck sees g as the appropriate measure of the
probability that a voter is in the relevant minimally sufficient set, and thus
a cause of the victory.
To take a simple arithmetic example, J gets 6000 votes out of a total of
9999 voters. So 3999 individuals voted for K. For J to win, J would have
needed 4000 votes: this is the minimal set of voters required to bring about
9
Tuck’s account faces complications, of course, in cases where votes are cast simulta-
neously, since the roll-call voting method of isolating which set of votes was, in fact,
sufficient, depends on temporal order. This fact plays an important role in motivating
Goldman’s account, which takes simultaneous voting as the normal case (on the grounds
that our voting conventions treat the timing of ballots, and of their being counted, as
irrelevant (p. 213)).
10
Writing about British parliamentary elections, Tuck notes that “at some point in the
course of the evening one candidate’s piles of ballot papers add up to the precise figure
necessary for a majority, and those ballot papers have therefore by themselves accom-
plished the task of electing him.” And he goes on to note that in the event of a recount,
the original count proves not to have settled the election and “the set which first reaches
the majority in the second count will then do so, or in the third count, or in however
many counts are necessary to satisfy the returning officer and the candidates that a proper
count has been made” (p. 43).
J’s victory. So each of the 6000 actual J voters has a two-thirds chance of
being in that set—and should on this basis be seen as having a two-thirds
chance of being causally responsible for J’s victory.
Goldman, in contrast, holds that in normal elections a vote is a “partial”
or a “contributory” cause as long as it is a member of any minimally sufficient
set, privileging no particular set as the one that, in the event, was s ufficient.
This means that every vote cast for a victor, even in a landslide, counts as
causally efficacious, not just those that are in some privileged minimally suf-
ficient set.11 As a result, the ex ante probability that one’s vote for a victorious
candidate will be causally efficacious, as Goldman thinks of things, is 1. (Of
course, when one does not know for certain which candidate will win, the
ex ante probability that one’s vote will be causally efficacious in bringing
about victory will be less than 1, reflecting the probability, whatever it is,
that another candidate will win, but in any case, still much higher than h.)
Importantly, in developing his view, Goldman focuses on cases of what
might be called simultaneous over-determination, where the sets of suffi-
cient conditions are all in place at the same time. In these cases it is hard, to
say the least, to make sense of the idea (important to Tuck) that one of the
minimally sufficient conditions takes precedence over all of the others. Yet,
Goldman recognizes the challenge posed by cases of non-simultaneous
over-determination (e.g. the roll-call vote and our national elections), where
the order of the votes, or the order in which they are counted, suggest that,
as a matter of fact, one sufficient condition pre-empts all the others. In
response, Goldman argues that in many elections we have a conventional
system in place that “abstracts from this actual or ‘natural’ order and consid-
ers all the votes on an equal basis” turning an apparent case of pre-emption
into one of simultaneous over-determination.12 We are unsure as to whether
the difference just noted is simply due to Tuck’s focusing on cases of non-
simultaneous over-determination, where one sufficient set of voters is rea-
sonably seen as pre-empting the others, and Goldman focusing on cases
of simultaneous over-determination, where there seems to be no grounds
for privileging one set of sufficient conditions over others.
11
Goldman advances this idea as a way of understanding Mackie’s INUS account of
causation in light of which something is a partial or contributory cause if and only if it is
“an insufficient but necessary part of a condition that is itself unnecessary but sufficient for
the result” (p. 206).
12
Goldman points out that “In the United States House or Senate, for example, a
roll-call vote is completed even if the outcome is clear long before the last vote has been
voiced. This is because, officially, votes are not counted or ‘registered’ until all have been
voiced. Because of this conventional feature, the causal impact of a late vote is not really
preempted by a collection of early votes. From the official, conventional perspective, they
are all simultaneous; hence, their causal statuses are perfectly symmetric” (p. 213).
It may be that each would accept the views of the other for the relevant
cases. What is important for us here is that Goldman and Tuck both reject
RCV, along with its emphasis on h, and recommend thinking of causation
in terms of the sufficient, rather than necessary, conditions for an outcome.
For the most part, differences between Goldman and Tuck (to the extent
there are differences) won’t matter to what follows, though where it does we
will default to Tuck, using g (interpreted as his view, rather than Goldman’s,
would recommend). This is for two reasons. The first is that we think that
in many cases (in elections and otherwise) the kind of convention Goldman
relies on in forestalling worries about pre-emption is not in place. And we
take it that when it is not we should be thinking in terms of pre-emption (as
per Tuck) rather than simultaneous over-determination. The second is that
the interpretation Tuck’s account gives to g makes that probability more
directly comparable to RCV’s h than is the interpretation provided by
Goldman’s account (which would set the probability at 1 in cases where one
is certain which candidate will win).
The contrast between g and h is striking. For a start, g is very much larger
in all democratically relevant cases. As noted in footnote 3, the probability
of an exact tie (h) between candidates in a US presidential election too close
to call is estimated at around 1/12,500. But for just such an election, g will
be quite close to 1. Even if the margin turned out to be 4 million voters out
of 120 million (perhaps not so very close after all) then g is 58/62 or roughly
0.94. On Tuck’s view, this probability measures the probability that one will
be a cause of J’s victory.
Note too that g and h vary in different ways. As electorate size increases,
the probability of an exact tie declines roughly with the square root of size:
so when M doubles, h declines by about 40%. By contrast, g is a matter of
relevant proportion, increasing as the size of the minimally sufficient set
increases.
It is worth emphasizing that g can be thought of as an ex post probability—
to be calculated after the election is settled—whereas h cannot be. After the
election, you will know whether you were pivotal or not. But the decision
as to whether to vote or not depends on the probability that one will be a
cause of the outcome in question; and so g should properly be thought of
here in ex ante expected terms. The ex ante expected value of g will be based
on the expected size of the majority. The voter properly expects h to be tiny;
but g is likely to be substantial.
In developing their accounts of how it is our votes might be causes of an
electoral victory without being required for it, Goldman and Tuck are espe-
cially interested in why voters “should vote.” They want to identify reasons
people have to vote even in cases in which one candidate will predictably
beat others, with or without their vote. Goldman explicitly recognizes that
the reasons in play might be either prudential or moral; and though his
own account focuses on the moral aspect, he thinks his reasoning has clear
implications for the prudential dimension as well. Specifically, Goldman
claims that:
the account [of causal responsibility] presented here . . . can explain
why people should vote (after obtaining sufficient information) and
it can explain why people do vote (in fairly substantial numbers)
[p. 216].
Taken in context, the idea, as we understand it, is that, once voters reckon
properly the probability of their vote being causally efficacious, they will see
that their vote is likely to have a causal impact on the outcome and will be
able to take some moral satisfaction, after the fact, in having been part of
the cause of their favored candidate’s victory. Furthermore, the prospect
of such moral satisfaction (and/or any social esteem that might attach to
recognition of their role as members of the group that brought about that
candidate’s victory) is a significant incentive.
Interestingly, Goldman and Tuck both talk not just in terms of causal
efficacy but also in terms of causal responsibility. And we shall have some-
thing to say later about the significance of this choice of terms. But we think
the account draws much of its intuitive force from its use of “responsibility”
language, since, with them, we do think that in elections all those who
voted for the victor—not just one pivotal voter—are responsible for the
outcome. So, consider, for example, an individual who voted for Hitler in
the German elections of July 1932. This individual, one might think, and
all the others who voted with her, must bear some responsibility for Hitler’s
successes in that election. Any such voter should properly be held in some
contempt; and she ought to feel guilty for having voted in this way.
Suppose, in her defense, she were to point out that her own individual
vote almost surely made no difference to the electoral outcome. Hitler, she
observes, would have been every bit as successful whether she had voted for
him or not. Would we be—should we be—moved by this observation?
Almost certainly not! Horribly bad moral consequences are associated with
her actions and she ought to take moral responsibility for them.13 And she
ought to do so specifically despite the fact that her individual vote almost
surely made no difference to the electoral outcome.
13
We leave to one side the important question of just how to think of the relative degree
to which she, as compared to others, is responsible for those consequences. We also leave to
one side questions concerning the responsibility she bears for her vote, regardless of the
consequences associated with it. (Here, we are being careful to talk of the consequences
associated with her vote, rather than of the consequences of her vote, since the latter assumes
her vote had the consequences in questions—that it was causally efficacious, and it is that
assumption that we are exploring.)
As will become clear below, we think Goldman and Tuck are right that
there is an important sense in which in elections a group of voters, not just
one, is causally responsible for the outcome. Our concern is with how
thinking about what one might cause figures in justifying and explaining
how people vote. In particular, we want to focus on two questions:
1. What exactly is each J-voter responsible for?
2. And does such responsibility that arises by virtue of being causally
efficacious (or causally responsible) exhaust electoral responsibility
more generally? For example, what account of voter responsibility can
RCV, as we have conceived it, put in play against the Goldman/
Tuck account—and does the best RCV account serve to satisfy our
intuitions in cases like the Hitler-success case?
We shall address those questions in turn in what follows. The upshots of
our discussion are twofold. The first is that there are serious reasons to
doubt whether the Goldman/Tuck account of why people should vote suc-
ceeds in giving the right action-guiding advice. In that sense, the normative
strand of the Goldman/Tuck account is suspect. Our second claim is that
the explanatory version of the account depends on assumptions that the
circumstances of voting make rather dubious; and that under plausible con-
ditions, the effect of the “Goldman/Tuck account of causal responsibility”
on anyone convinced by it may well lead them to be less likely rather than
more likely to vote. In that sense, the explanatory strand of the Goldman/
Tuck argument is suspect.
Both RCV and the rival Goldman/Tuck account are about choices:
the choices individuals do make; and the choices they ought to make. In the
choice context, the proper weight of various considerations bears; and, in
this context, we think two questions loom: “What value is to be placed on
being the cause of some outcome?” and “What probability is to be assigned
to that value?”
The stakes for RCV orthodoxy are considerable. Given that g can be
close to 100% for close elections (and far from negligible for even not so
close ones):
1. We should expect, contra RCV, that the high probability that one’s
vote will be causally efficacious will motivate many people to expend
time and energy in voting. This is, of course, precisely the point that
Goldman and Tuck seek to make.
2. We should correspondingly expect voters to expend significant
effort in acquiring relevant information about candidates and
their policies, contra Downs and Caplan. (Unless, of course, they,
perhaps mistakenly, think that they already have the relevant
information.)
3. We should expect that considerations of individual self-interest

are likely to play nearly as large a role in electoral settings as in those
same agents’ market decisions. After all, the probability that one will
be causally efficacious (on the Goldman/Tuck account) will often be
nearly the same in both contexts.
One crucial question is whether g or h is the way that people actually
think about the probability that their vote will be causally efficacious.14 The
difference will have a significant impact when it comes to explaining voting
behavior. Another crucial question is whether g or h is the appropriate way
to think about the probability that one’s vote will be causally efficacious.
The difference will have a significant impact when it comes to justifying
voting behavior.
2. RESPONSIBILITY FOR WHAT?
Goldman and Tuck tie their account of moral responsibility to the notion
of causal efficacy. We think that this elision is misleading. We think a proper
account of moral responsibility is often usefully kept apart from any particu-
lar account of causation, and, indeed, in many cases, apart from causation
altogether; and that this is true in the electoral case. Several considerations
incline us to this view.
For one thing, we think that the Goldman/Tuck formulation, by
grounding the idea of moral responsibility in its causal effects on the elec-
toral outcome, fixes the idea of moral responsibility in electoral behavior
too narrowly. It is not that we think that bringing about an electoral out-
come is totally irrelevant to the ethics of voting behavior, but as we shall
argue, it is not the only relevant aspect. More generally, though both respon-
sibility and causality are notions whose application turns on context and
convention, the relevant contexts and conventions are significantly different
in the two cases.
Take causation. It seems clear that in most (perhaps almost all) possible
descriptions of what caused some ‘outcome’ X, there are a number of factors
that might be isolated all of which were present—and necessarily—for X to
14
We believe it is likely that many people do think that as long as they voted for the
winner of an election, they helped to cause the victory, as Goldman’s (but not Tuck’s)
version of g would have it. Yet, we suspect that thought does not depend on ideas con-
cerning their being in a minimally sufficient set. In any case, it does seem plausible that
many people recognize that they cannot claim credit for a victory without voting and
suppose that as long as they do vote (for the winner) they can claim credit. And we think
that is likely a real incentive to many people.
come about. But what counts as a cause of that outcome—and what in

particular as “the” cause, or a primary cause—can be highly contextual. As
Goldman puts it, “causal upshots” can be “defined or stipulated by social
convention.”
Nevertheless, causation does not seem to be as permissive in this respect
as is responsibility. Take the linguistic indicators. Common parlance allows
variously: “holding responsible”; “feeling responsible”; “taking responsibil-
ity”; “being responsible”; “acting responsibly”; and related expressions. All
these states of feeling/holding/taking etc. refer to social and psychological
facts of an impressively wide variety. When people talk of their “taking
responsibility for X,” they seem to do so not so much because they believe
that they have caused X but rather in face of the conviction that they have
not. Equally, when people take the trouble to declare that they are holding
A responsible for X, it is often enough in a setting where the causal connec-
tion between A’s actions and X is either unclear—or seen as irrelevant. The
analogous linguistic indicators and practices with respect to causation seem
different and much more constrained. People do not talk so readily of
“accepting causation”—except perhaps in the rather loose sense in which
causation might be connected to responsibility. We might invite others to
treat us as if we were causally efficacious but only as a maneuver in getting
them to assign responsibility to us: causal efficacy remains “as if ”—while
the responsibility we take on, we do indeed treat as ours.
It is, we think, unexceptionable that when A votes for a winning option
J, people at large are disposed to “hold A partially responsible for J’s victory”
in a way that they would not hold B, who abstained, responsible; and in
turn more than C who voted for K. But this is a weak claim—entirely con-
sistent with the thought that A’s voting for J rather than for K (or abstain-
ing) almost surely made no difference to whether J would win. A’s
responsibility might derive from the fact of A’s supporting some outcome,
even if, as we think, it is not restricted to A’s being causally efficacious in
bringing the outcome about.
Consider, for example, citizen A. Suppose that J and K are rivals in an
electoral race. Suppose that it would be desirable from a moral point of view
if J were to win; and that A recognizes this fact. So, A thinks, in any case
where the chances that J will win are pretty high, that fact is a good thing
(better than if J were likely to lose). A has, as we might put it, a morally
grounded “pro-attitude” towards J’s victory. In that attitudinal sense, A sup-
ports J.
However, A could support J in another way—by bending her own agency
to the project of J’s election by voting for her. In that way, so the thought
goes, A might well turn out to be a cause of J’s victory in a way that A
wouldn’t be if she failed to vote.15 And when J is duly elected, A could

properly feel moral satisfaction for her own part in supporting this desira-
ble outcome.
Now, it is important to emphasize that A’s contribution here does not
collapse to the difference A’s vote made to whether the good outcome would
come about. That effect is fully captured by the h factor—which by hypoth-
esis here is very small; J’s chances of winning without A’s “contribution” in
the postulated example are already high. Perhaps A’s “responsibility” lies, as
the Goldman/Tuck account would have it, in A being, as they think about
it, a cause of the victory. But we think that does not capture all that is in
play, and, in fact, we suspect it misses the most important part: that what
A is doing when he votes for J is exercising agency in support of the relevant
outcome. And that seems to us to be the appropriate turn of phrase.
Accordingly, we want to say a little more about what is at stake in this
“agency” thought.
As the Goldman/Tuck view in effect emphasizes, in lots of situations,
individuals want not just that something good comes about but also that it
is they who bring it about. A wants his partner to enjoy pleasure: but it is
likely that A wants that pleasure to be at A’s hands. We ourselves desire
intellectual progress in understanding “reasons to vote”—but we also want
that progress to come about partly as a result of our work—not just that
some paper be written and make its contribution, but that it be our paper.
In a similar way, it is a familiar experience that a newly appointed CEO (or
new dean in the university setting) wants to make changes in the way things
are done, or in the personnel who are her underlings—not just because
those changes are good in themselves (though she may well think so) but
also because she wants to leave her mark. She has the power to make changes
for the better and she relishes the exercise of that power.16
In the case of the new CEO/dean (and perhaps the others as well), this
motivation can appear to include an “ego trip” element: she may be making
the changes simply because she can, and largely independent of whether the
changes represent genuine improvements. But this (limiting) case reminds
us that the exercise of agency can come apart from approval of the changes in
the outcome qua outcome. And once we recognize that the exercise of agency
15
Whether she is a part of the cause depends, of course, on whether J wins and also,
on the Goldman/Tuck account, on whether her vote was part of (on Tuck’s version) the
minimal set of votes that was sufficient for J’s victory or (on Goldman’s version) a mini-
mal set of votes sufficient for J’s victory.
16
As Tuck notes acknowledging a concern with agency, and with being oneself the
cause of an outcome, is perfectly compatible with a holding that RCV offers the right
account of when a prospective outcome provides instrumental reason to act. See pp. 54–7.
can be desired for its own sake, we can see that changes might be made even
when the agent does not know or much care whether such changes would be
desirable were they to come about in some other way. A person may think:
“these are my changes and I am making my mark upon the world.” And that
exercise of agency may well be the source of her satisfaction.
Note that, since on the Goldman/Tuck account the reason to vote relates
to being causally responsible for the outcome, the desire to be a cause pro-
vides individuals with incentives to vote for the candidate most likely to
win. This is so because on the Goldman/Tuck account, you cannot be caus-
ally responsible for something that doesn’t happen: voting for a losing can-
didate deprives you of anything to be responsible for.17 So, if you want to be
causally responsible for an event in political history, just for the sake of
leaving your mark, you need to vote for the winner.18
It is worth noting too in passing that the notion of agency at stake here is
distinct from that implicit in most rational actor theory, because in RCT
any and all rational action is seen as the outcome of choice. So any choice is
an example of the individual exercising “agency.” In particular, the choice to
abstain rather than to vote is a choice—and hence abstaining becomes an
action, an instance of agency. By contrast, the notion of “agency” we see as
being brought to the fore in the Goldman/Tuck account of voting has a
more common-sense quality: to play tennis or go jogging is to exercise
agency in a way that choosing to do nothing would not be; to choose to
intervene in the world is to exercise agency where to choose to let things
take their course would not be. And specifically, to vote is to be active in a
process, in a manner that abstaining would be passive. Both voting and
abstaining involve choices; but only one of those choices involves one’s full
political agency. Or at least, so the thought goes. And it is by no means an
unfamiliar thought. The distinction between killing and letting die depends
precisely on a notion of agency in this “common sense” sense. The decision
as to what to do in familiar trolley problems involves a choice; but “agency”
is involved only when you pull levers or push fat men.
Of course, in the CEO/dean cases, and many others, the agent is fully
causally efficacious in bringing about the outcomes that represent her mark
upon the world. That is not so in the voting case. Nevertheless, there can be
17
“On my account,” Tuck writes, “it is rational to vote (all other things being equal)
only if I believe that there are likely to be enough votes for my candidate for my vote to
be part of a causally efficacious set . . . In other words, I think that it is precisely in the
situation where it looks on the standard modern view [RCV] as if my vote is unnecessary
that I have a good reason to vote” (p. 60).
18
For an account of voting behavior based on the related, but importantly different,
conjecture that voters want to be on the winning side for its own sake (at least over some
range), see Schuessler (2000).
an agency effect—something for the voter to be responsible for, over and

above that voter’s (usually very small) impact on the likelihood of J’s victory.
A has chosen to be causally involved by bending her agency in the further-
ance of J’s election; and the spirit of the Goldman/Tuck account seems to be
that there is a level of causal responsibility associated with the fact of “being
a cause”—distinct from her individual contribution to likelihoods. Further,
this agency effect can both: (a) motivate individuals to vote; and (b) provide
a moral justification for A’s voting.
The (a) aspect we find perfectly plausible. Individuals might well be
induced to vote for candidates because they want themselves to be a cause of
that candidate’s victory. In what follows we grant this point.19 At the same
time, we note that where people do have this concern, their voting has a
very high probability of being pivotal to their success in pursuing that end.
However, the (b) aspect requires more argument. We think there is at
least a range of cases in which including the agency effect will secure a worse
outcome than if no such effect were present. In that sense, the Goldman/
Tuck account of causal efficacy seems to give the wrong action-guiding
advice. To show this is the aim of the ensuing section.
3. ACTION-GUIDING ADVICE
Begin with a simple example. There is a community of voters and an elec-

tion in which the moral stakes are seriously large. The expected outcome is
that 60% of voters will vote for J. This means that the probability is quite
high that in voting for J one will be (in the Goldman/Tuck sense) a cause of
J’s victory, and so able properly to see oneself as having brought it about
(along with others).
Now consider, in this setting, a specific voter A, who has an opportunity
to bring a real, but not especially large, benefit, of E, to the community.
There is, however, a difficulty—namely, that securing this benefit requires
A to be out of town on voting day and hence unable to cast a vote. Should
A remain in town and vote; or secure the benefit of E? On the stipulated
assumptions, we think it is clear: A should go out of town and secure the
benefit, even if the value of J’s victory is much greater than the value of E.20
This is because it is virtually certain that J will win (since 60% of the electorate
Though we will have cause to revisit that issue briefly in section 4.

19
Of course, there is some difference in the relative values of J’s victory and of E that
20
would recommend staying and voting. But for a broad range, the fact that J’s victory is
more valuable will in effect be irrelevant to what A has reason to do, given the expected
votes of others.
is expected to vote for J): A’s vote will almost surely not make a difference to
the outcome. That is, h is very small. At the same time, g is substantial and
the Goldman/Tuck emphasis on whether one will be a cause (in their sense)
of a valuable outcome misdirects attention to the possibility of being a cause
of that outcome, ignoring the crucial fact that one is close to decisive over E.
Absent there being an independent (and substantial) value to being a cause
(again, in the Goldman/Tuck sense) of some outcome, a concern for the
value of outcomes ought to lead one to be both indifferent to g and attentive
to h in determining what one should do.
Goldman and Tuck may dispute that A can know with total certainty
that J will win. We, of course, concede that point. Suppose we allow that
A’s absence does reduce the probability that J will win by a certain frac-
tion. So the expected cost of A going out of town is not zero after all.
That does not bring the Goldman/Tuck account of causal responsibility
in line with reasonable action guiding advice, unless the probability that
J might lose is extremely high (given the assumed values at stake). And to
insist that it matters that the probability is not zero is, in effect, to con-
cede what is at stake—namely that what is relevant for determining what
A should do is not the share of so-called “causal responsibility” (as
Goldman/Tuck would have it)—not the probability that A, if she votes,
would have been one of the people that counted to make up the minimal
majority—but rather the likelihood that J will lose because A doesn’t vote.
In short, h.21
The example could be set out, not in terms of voting, but in other terms
where the same issues are in play. Suppose the Coast Guard has two boats—a
small one and a large one. The small one requires only a one-person crew;
the large one requires a six-person crew. There are seven people on call at the
station and two distress calls come in from different points along the coast.
One involves a ship with 140 people; the other involves a boat that has just
one person. All seven could attend to the ship with 140. Or six could crew
the larger boat, while the seventh takes the smaller boat out to save one
more. The best outcome is achieved, we assume, if both boats go out and
141 are saved. We take it that under these circumstances the right action
guiding advice is that someone take the small boat out.
But it is far from clear that this outcome is what the Goldman/Tuck
approach would recommend. Consider the calculus of each of the coast
21
Tuck acknowledges that “it might obviously be a good reason for doing something
that I am probably the only person able to do it” (p. 59), so it may be that our disagree-
ment is more a matter of emphasis, than principle. But we are struck by the extent to
which Tuck (and Goldman) think the relevant consideration in contexts of voting is
whether one will likely be part of a relevant set of votes that is minimally sufficient for
victory, and not whether one’s vote is likely to be necessary.
guards. Each can go with the larger ship and take moral credit for saving 140.
(Just how much credit redounds to each is unclear. One reasonable sugges-
tion is that the credit should be divided equally among all who participate
in the rescue, or among the six in the minimally sufficient set; alternatively,
though, it might be (as Tuck argues) that each rescuer in the minimally
sufficient set can take full credit for saving all 140,22 still another is that the
credit to be shared equally is determined by the size of the minimally suffi-
cient crew.) Or one could crew the smaller rescue-vessel (receiving full
moral credit for saving just one person) and allow her colleagues to crew the
larger ship.
Suppose for the example that, out of gratitude, each person saved pro-
vides a reward of $N to the crew that saves her, and let’s tentatively use that
reward as a proxy either for the moral credit, or for the value, of saving
them. Then the crew of the larger boat will collectively receive $140N (to
be shared in some way by those who are causally efficacious), whereas crew-
ing the smaller boat promises only $N. Absent side payments—and irre-
spective of whether each in the larger group divides the money, credit, or
utility, evenly, or each can claim the total amount of what they cause (as
understood by Goldman and Tuck)—each reasonably prefers to crew the
larger boat. And this remains the case whether the reward comes as money,
or social esteem, or moral satisfaction.23 Of course, there is a reward/credit-
sharing scheme that will create incentives to produce the best outcome
(which, in this case, we take to involve maximizing lives saved). That scheme
will require that the total amount of reward to be distributed across all seven
in a way that is insensitive to who saves whom. This will secure the best
outcome because, when the seventh goes with the larger boat rather than
the smaller, her participation serves to reduce pari passu the reward each of
the six would otherwise receive. The total reward (monetary, social, or
moral) across all who crew would be maximized when all 141 are saved.
And one can certainly imagine a “reward-sharing” scheme that would secure
this result. But note that that “best-outcome” reward-sharing scheme is more
difficult to imagine operating in the case where the reward comes in the form
of esteem or moral satisfaction. Monetary rewards are readily transferable
22
Tuck focuses on the allocation of utility, not moral credit, arguing that in an elec-
tion that has an outcome with a utility of 200, each vote that is part of the causally suffi-
cient set, “represents” a utility of 200. But he cautions that we shouldn’t conclude “that it
possesses such utility, precisely because the notion of representation implies that the
thing represented is different from its representative” (pp. 42–3).
23
The difference between h and g makes all the difference here, since, however, the
credit or utility is to be apportioned, the much higher value of g, compared with h, makes
the expected utility of joining the crew of the larger boat much greater than the value of
crewing the smaller boat.
between persons: esteem and moral satisfaction are not. That is, it seems
plausible to suppose that public esteem and even “moral self-satisfaction”
associated with worthy acts (like saving people in distress) accrues to those
actors who are “causally responsible” whether in whole or in part. Securing
the “incentive-compatible” reward-sharing would require a separation of
reward from agency that seems entirely alien to the spirit of the Goldman/
Tuck treatment.24 The divergence between what action serves to secure the
best outcome and what action serves (on the Goldman/Tuck account) to
maximize the individual actor’s moral credit is troubling. It is troubling not
least because, for Goldman and Tuck, the analysis is supposed to explain
why individuals will have incentives to behave in particular ways—and if
moral credit is indeed allocated according to the Goldman/Tuck scheme,
then it is clear that the outcomes secured will often not be best.
There are really two issues here, as we see it. One concerns which scheme
would provide incentives that would predictably provide better outcomes.
The other concerns which way of thinking about what one might cause by
acting provides the right account of what one has reason to do. (We here
take no stand on how tightly connected these two concerns are, though we
think they are connected.) Our main point is that the weight the Goldman/
Tuck account gives to being (in their sense) a cause of an outcome gives the
wrong action guiding advice, if what matters is the moral (or other) value of
the outcome.
We recognize that these examples may not be decisive. There may be
other examples where the Goldman/Tuck approach might produce better
outcomes than the RCV equivalent. Furthermore, we concede that quality
of outcomes qua outcomes may not exhaust the moral domain. Nevertheless,
it seems to us that failure to provide the right action-guiding advice in
the case we describe (and the many others that have the same structure) is a
significant count against the Goldman/Tuck account. To meet this charge
defenders of the Goldman/Tuck account need arguments for thinking
either that each person should indeed join the larger boat or that their
account does not make that recommendation. As far as we can see, Goldman
and Tuck provide no such arguments. Nor do we see how those might go.
24
One might seek to block the force of the Coast Guard example in a number of ways.
One might be to insist that the seventh is likely not a cause because she adds nothing to
whether the project of saving will be successful, assuming the other six are in place. This
is a point defenders of RCV would make (even as they allow that there is a small chance
that the seventh would prove necessary to success since, for instance, one of the others
might become sick or fall over-board). But clearly, since only six are needed, and so only
six are in the minimally sufficient set, g, for each, will still be extremely high on Tuck’s
interpretation (and on Goldman’s, since each of the seven is a member of some set of
sailors that is sufficient).
4. ELECTORAL RESPONSIBILITY WITHIN

EXPRESSIVE THEORY
Nothing that we have said so far denies that when A votes for J, A is in fact
supporting J’s victory, nor is it to deny that A might have a high probability
of being (in the Goldman/Tuck sense) a cause of J’s victory. What we deny
is that that probability (g) provides much by way of a reason to vote. But
that doesn’t mean that J has no substantial reason to vote. Indeed, we think
people find, and are right to find, significant value in exercising their agency,
in no small part as a way of expressing their political commitments. And we
think people are rightly seen as responsible for exercising their agency in this
way (as in others). Yet, the importance of exercising agency is in play regard-
less of whether one votes for the victor or not. What matters in elections, we
are thinking, is both the outcome and the exercise of agency, but not that in
exercising that agency one happens to be among those who (in the Goldman/
Tuck sense) cause the outcome.
When A votes for J, she does three things: first, she reveals certain things
about her attitudes and beliefs; second, she reveals her preparedness to express
those attitudes and beliefs; and third, she expresses those beliefs specifically
at the ballot box. All of these attributes/actions are ones for which A can
properly be held responsible (and are proper grounds for pride or guilt). The
moral responsibility for those attitudes and expressions applies, over and
above any causal influence on J’s election and, indeed, independently of
whether J wins.
Consider attitudes first. People can be, and typically are, held responsible
for their political attitudes, whether or not those attitudes actually bring
about the states of affairs that are the content of those attitudes. Someone
who holds the view that Hitler’s eugenics policies were admirable surely
does deserve our contempt on that account, even if she never does anything
in relation to those policies other than to admire them. If A’s moral judg-
ments are defective then A is morally defective to that extent.
But expression of such attitudes involves a further step. Sometimes, when
people have odious views, it is best that those people keep such views to them-
selves. They ought to be silent at dinner parties when certain topics come up;
they ought to refrain from writing op-ed pages; or calling in to live radio
shows. It is one thing to hold the views in question, another to express them. If
A gives expression to the attitudes in question—if he declares them—he is
endorsing them, giving them a public life that they do not deserve.
We are drawn to the Goldman/Tuck idea (as we interpret it) that when A
expresses his attitudes at the ballot box he is expressing them in a distinctive
way—in a way that may have causal upshots, however small. For given the
effects on the probability of success of his favored candidate are positive, A

can be thought of as “bending his agency” to the cause of that candidate’s
victory, whether or not he is a cause of that victory. Voting falls under the
description of “agency-bending” because the consequential effects are strictly
non-zero. But it would be a mistake to think that because the agency effect
depends on h being non-zero, the agency value is a direct function of the
degree of causal impact. A’s bending his agency towards the cause of J’s vic-
tory has to be sharply distinguished from the causal impact of A’s vote on
the likelihood of J’s victory.
So when A votes he can be held morally responsible for the political atti-
tude that vote reveals, for expressing the attitude, and for the mobilization of
his agency in a given direction. Note that these are all things that depend
not at all on what any other voter does, nor on what the outcome of the
election ends up being, nor on whether one is among those in the set of
votes that were minimally sufficient for victory.
But here is one (further) sense in which this “expressive” version differs
from the Goldman/Tuck account. A central feature of the Goldman/Tuck
account is that what individual voters are (partially) responsible for is the
electoral outcome, and hence that only those voters who vote for the actual
outcome can bear this responsibility: voters who vote for the unsuccessful
candidate have no outcome to be responsible for. But on our view, those
who vote for a bad, but happily unsuccessful, candidate do not avoid
responsibility just because their candidate was unsuccessful. Bad but unsuc-
cessful voters bear the responsibility for their odious attitudes, for their
expressions of those attitudes, for bending their agency in a bad cause, and
for the risky behavior they undertook in promoting the chances of a seri-
ously bad outcome. And these components of moral responsibility remain,
whether the candidate won or not. Of course, Goldman and Tuck need not
deny this. They do not claim that their account of causal responsibility
exhausts all that might be said about electoral responsibility. However,
when these other aspects of responsibility are included—elements that the
RCV account can in principle include—it is not so clear that the Goldman/
Tuck account of specifically causal responsibility adds much.25
In summary, a plausible account of what one is morally responsible for in
voting provides a notion of responsibility that captures what lends the Goldman/
25
And we are concerned that taking g, rather than h, as the probability relevant in
making choices obscures, rather than clarifies, what is at stake. When it is important to
an agent to exercise her agency, as we recognize it might well be, we think it appropriate
to use h, not g, in thinking about the expected value of the options (in these cases not just
the agency-independent value of the potential outcome of that exercise, but also the value
of exercising agency). Of course, choosing to exercise one’s agency is, normally, both
necessary and sufficient for doing so, so h, in such cases, will be 1.
Tuck view most of its intuitive force. Such an account suggests that a voters’
causal efficacy in relation to electoral outcomes is rather less important than
Goldman and Tuck seem to imply. Certainly, on the account of responsi-
bility we’ve highlighted, individuals will systematically be morally responsible
for their beliefs and attitudes, and their expressions of them, and how they
cast their votes, and will remain so whether they vote for the winner or the
loser. At the same time, we suspect that, if one has bent one’s agency in
support of an outcome, and it comes about, one is to some degree respon-
sible for that outcome—whether or not one was (in the Goldman/Tuck
sense) causally efficacious in bringing it about.
5. THE NORMATIVE VERSUS THE EXPLANATORY
The Goldman/Tuck account purports to show not only why agents ought
(or at least have moral or prudential reason) to vote; but also to explain why
significant numbers of them actually do. We have already indicated why the
normative aspect of the argument is questionable. But as we indicated at the
close of section 2, the explanatory and the normative aspects of the account
are less closely linked than Goldman and Tuck suppose. And, indeed, as we
indicated there, the desire to exercise agency in and of itself might well
provide a plausible motive for voting. So, insofar as that desire is what
Goldman and Tuck have in mind by “causal responsibility,” the explanatory
part of the argument seems to proceed intact.
To be sure, that explanatory story requires certain assumptions about the
role of moral factors in agent motivation—the idea specifically that agents
are motivated in part by the prospect of moral satisfaction and social esteem.
These motivational assumptions might be controversial in some RCV cir-
cles but they are ones we broadly endorse, and so we shall not discuss them
further here.
Nevertheless, to the extent that these motivational matters are in play,
it is not entirely clear that they do indeed give rise to higher turnout
than would arise if agents based their voting decisions on the standard
RCV calculus—or some other calculus in which perceived causal efficacy
in producing the electoral outcome to be lower. To give a sense of our
misgivings here, return to the example of agency effects involving the
new dean/CEO, exercising her agency largely for its own sake. She may
well receive satisfaction from “leaving her mark” and she may do so
somewhat independently of whether the mark so left is actually desira-
ble or not. Perhaps she thinks her innovations probably are consequen-
tially desirable; but it may well be that her epistemic warrant for this
belief is somewhat tenuous. But now she reads Goldman and/or Tuck and
realizes that she is more accountable for her actions than she had previously
realized. She now sees that the moral and reputational stakes are higher than
she had thought: her “sense of responsibility” increases.
Accordingly, if there is a serious possibility that the changes she intended
to make may be changes for the worse, then she may well be inhibited in
making them. In the same way, if there is a serious chance that J will turn
out to be an inferior candidate, any increase in perceived responsibility may
well inhibit A from voting. The Goldman/Tuck examples take it as given
that J’s election is independently morally desirable.26 But that assumption
occludes real-life uncertainty about the moral qualities of candidates and
their policies.
Recall that by hypothesis there is a significant vote for K as well as J. Of
course, motives for voting can be various; but in the Goldman/Tuck spirit,
it seems reasonable to think that many of those K voters actually believe
that K, not J, is the morally superior candidate. In the absence of any con-
trary argument, this fact should give the J-voter pause. Confidence in the
proposition that J is indeed superior ought to be somewhat shaken.
And in the face of increased uncertainty about the moral qualities of
options, A’s voting calculus takes on a different hue. After all, we take it that
it is a worse thing to bend your agency to the cause of the worse candidate
than it is to abstain. So the effect of increased uncertainty about whether J
is indeed the superior candidate to K seems bound to reduce turnout. And
the effect of increased responsibility, in the face of a given level of such
uncertainty, seems likely to have a similar effect. Greater probability that
one will cause an outcome—a greater chance of being causally responsible
for the outcome—will encourage individuals to acquire more information
about candidates given that they intend to vote. But whether, given the cost
of information acquisition and the risk of making an error, the greater prob-
ability of responsibility will also induce more individuals to vote seems at
best an open question.27
It is worth noting that increased perceived probability of responsibility
may lead A to abstain even when A is certain that J is the superior candi-
date. If what morally commends J is not that J is good but rather that
26
Goldman considers cases in which someone’s preferred candidate is objectively
worse and argues that when that is true the person does not have (an objective) reason
to vote. (He notes that she may still have a subjective reason, depending on the evi-
dence she has concerning which candidate is better.) But he does not consider, as we
do in what follows, the impact the recognition of the possibility might have on voter
behavior, in light of the more expansive account of causal responsibility he and Tuck
defend (pp. 209–10).
27
What is at stake here is how the moral payoffs to the three prospects—voting for the
better candidate, abstaining, and voting for the worse candidate—each respond to
changes in the perceived likelihood of responsibility.
K is even worse, A may have to wear a certain social and moral oppro-
brium just for voting for a bad candidate. In such a case, increased per-
ceived probability of responsibility for the outcome may leave A wanting
to keep her hands clean—abstention may here too emerge as the pre-
ferred option.28
What greater perceived responsibility will do is to increase the propor-
tion of voters who are morally confident or who are ignorant as to how
ignorant they actually are. Voting is left not only to those who properly see
themselves to be relatively well informed, but also to the opinionated, the
self-deceptive, and those who vote for non-morally grounded reasons.
A heightened sense of responsibility, as Goldman and Tuck think a more
proper view of causal efficacy would generate, will not necessarily lead to
higher turnout. But it probably will lead to a voting body that is more mor-
ally confident, whether because voters have gathered more information or
because of independent psychological factors (whose normative status
seems rather more dubious).
6. CONCLUSION
The “causal responsibility” account of voting behavior, advanced by Goldman

(1999) and endorsed with some variation by Tuck (2008), purports to show
why individuals should vote, and why they will vote in larger numbers than
they would if they were informed by the standard rational choice account of
voting behavior. According to the latter account, the ex ante probability of
influencing the electoral outcome is the probability of a pivotal result—one
in which voters for the successful candidate are individually necessary and
jointly sufficient for the electoral outcome.
In our assessment of the Goldman/Tuck account, we have wanted to
attend to three issues:
1. What is the distinction (if Goldman and Tuck see one) between
causal efficacy and causal responsibility?
2. When an individual causes, or is a “part of the cause,” of a
candidate’s success, what is she responsible for? Is she responsible (at
least in part) for the electoral outcome, or for being (in the Goldman/
Tuck sense) a cause of her favored candidate’s success? If the latter,
is there a material difference between being a cause of the candidate’s
success and “bending one’s agency towards (the cause of ) that
candidate’s success” (which is our own preferred way of thinking
about such cases)?
28
We are grateful to Emma Johnson for this observation.
To the extent that the issue does revolve around agency effects so under-
stood, two further questions arise:
a. Does the inclusion of agency effects give rise to the correct ‘action-
guiding’ advice in electoral contexts?
b. Is it plausible that including such effects will encourage increased
turnout in the circumstances of normal elections?
Our answers to these two questions are negative.
Perhaps a negative response in relation to the first should occasion no
surprise. Agency effects invoke factors that seem more at home in a ‘virtue’
account of electoral behavior than a standard consequentialist one; and it
ought to be expected that adding normative requirements of this voter-
virtue kind might ‘cost’ something in terms of the quality of outcomes qua
outcomes. To be sure, Goldman and Tuck do not cast their “causal respon-
sibility” story in virtue language; but we suspect that that is where their
account properly belongs.
As to the second, explanatory question, the issue hangs on whether see-
ing oneself as more rather than less likely to be responsible for the outcomes
of one’s voting behavior increases the incentive to vote. Much depends on
the weight of guilt/shame when one votes for the “wrong” candidate vis-à-
vis moral satisfaction/pride when one votes the right one. If the latter is less
weighty than the former (as we think most likely) then the best course, in
the face of uncertainty about which is the better candidate, may well be to
abstain. Goldman and Tuck steer clear of such uncertainty in their examples—
but it is difficult to ignore the prospect of this kind of error in fact, or
indeed to account for the patterns of voting behavior that Goldman and
Tuck assume in their examples (and especially for anything other than land-
slide victories) without risk of voter error being part of the story. Of course,
having voters abstain when they would otherwise have voted for the worse
candidate is presumably a good thing. But the expected value of that effect
is properly measured by relying on the parameter that RCV theorists have
long insisted is the proper parameter—namely, h.
Bibliography
Brennan, G. and L. Lomasky (1993). Democracy and Decision, New York: Cambridge
University Press.
Brennan, G. and P. Pettit (1990). “Unveiling the Vote,” British Journal of Political
Science, 20: 311–33.
Caplan, B. (2007). The Myth of the Rational Voter, Princeton: Princeton University
Press.
Downs, A. (1957). An Economic Theory of Democracy, New York: Harper.

Goldman, A. (1999). “Why Citizens Should Vote: A Causal Responsibility
Account,” Social Philosophy and Policy 16: 201–17.
Riker, W. and P. Ordeshook (1968). “A Theory of the Calculus of Voting,” American
Political Science Review 62: 25–42.
Schuessler, A. (2000). A Logic of Expressive Choice, Princeton: Princeton University
Press.
Tuck, R. (2008). Free Riding, Cambridge, Mass: Harvard University Press.
Tullock, G. (1967). Toward a Mathematics of Politics, Ann Arbor: University of
Michigan Press.
PA RT I I
POLITICAL LIBERALISM
AND PUBLIC REASON
3
Political Liberalism
Its Motivations and Goals
Charles Larmore
In recent years, political liberalism, understood as a distinct form of liberal

theory importantly different from the classical liberalism of such great
thinkers as Locke, Kant, and Mill, has become a vigorous area of philosoph-
ical reflection. The amount written on it, both expository and critical, is
now enormous. Having helped along with John Rawls to introduce this
conception,1 I am pleased by its vitality. But I also see matter for concern.
For as political liberalism comes to form yet another industrial site in the
great business of academic philosophy, its distinctive motivations and goals
as well as the way it continues, while deepening, the cause of classical liber-
alism may be easily lost from sight. This danger I mean to counter in the
present chapter, laying out, as I see them, the problem to which political
liberalism aims to provide a solution, the means—in particular, the moral
assumptions—by which it seeks to solve its problem, and the ends it can
reasonably hope to achieve by this solution.
1. ORIGINS
I begin by explaining why liberalism in general, by its nature and not merely
by historical accident, is a latecomer among the forms of political life.2 In the
past, political association, seeking as it must always do to unite by means of
binding rules a group of people for the achievement of collective goals, first
took the form of communities organized around some single, authoritative,
more or less regimented definition of the human good. Only if people are
at one in their own ultimate purposes, so it was understandably supposed,
1
Larmore (1987); Rawls (1996).
2
Cf. the discussion of this point in Larmore (1996a), 142–4, 211–13.
64 Charles Larmore
will there be the social cohesion necessary to achieve the purposes of the
whole. In pre-modern times, the dominant ideas of the ends of existence
were generally religious in character since the world itself was seen more
as a realm of superhuman powers to be propitiated than as an object of
prediction and control. At the same time, those in positions of power were
only too eager to encourage such ideas for the more-than-human author-
ity they could themselves draw from them. Politics was thus religious, and
religion political. To be sure, reigning views of the human good did not
go uncontested. But few people believed that society is possible except on
the basis of some shared and even sacred definition of the meaning of life.
As Lactantius memorably summed up the dogma for the Christian Europe
of late antiquity and the middle ages, “it is the fear of God alone that
secures the mutual society of men, by which life itself is sustained, protected,
and governed.”3
Why this common understanding should have begun to crumble in the
16th and 17th centuries in Europe is a complicated story. No doubt, medie-
val developments—the constant tensions between church and state (sacerdo-
tium and imperium), the succession of reform movements from the 10th
century on that sought to recover the true Christian teaching and gave
increasing importance to inner faith and conscience—played a large part.
Whatever the full explanation, early modern times saw the ever more wide-
spread realization that people reasoning sincerely and carefully about ques-
tions of faith and about the nature of the human good in general are likely
to disagree, often because they differ about what it means in these cases to
reason well. To the long-standing problems of political rule—how to tame
the passions and settle conflicts among interests in order to secure the condi-
tions of social cooperation—was added the need to rethink the basis of polit-
ical authority itself, given the breadth of reasonable disagreement about the
ultimate ends of life. This problem is not one simply of people disagreeing
about religious and ethical questions and about their implications for the
organization of society. That too is an age-old difficulty. At issue was the
widening recognition that reasonable people, by their very exercise of reason,
tend to disagree about these matters. For therein lay a fundamental chal-
lenge to the cogency of any attempt to justify, as all political regimes must
seek to do, the system of rules by which people are to be governed.
Early modern times thus saw as well the emergence of various views
about how to handle this new problem—views, in particular, about how to
bypass the profound religious and ethical disagreements and find in such
factors as self-interest, a regard for the favorable opinion of others, or a sense
3
Lactantius (c. 240–320 ad), De ira Dei, XII: “Timor igitur Dei solus est, qui custodit
hominum inter se societatem, per quam vita ipsa sustinetur, munitur, gubernatur.”
Political Liberalism: Its Motivations and Goals 65
of what is right and fair the basis of the principles by which people can live
together peacefully and fruitfully. The problem and its solution were a pre-
occupation of such seminal figures as Bodin and Montaigne, Hobbes and
Bayle. They constitute as well the dominant concern of the liberal tradition
that begins paradigmatically with Locke (though the thinkers just men-
tioned count among its progenitors) and has undergone further key devel-
opments in Kant and Mill and to the present day. The central place in this
tradition of theories of toleration attests to the formative role played by the
effort to reconceive political society in the light of reasonable disagreement
about the makeup of the human good.
Now the cardinal principles of a liberal society—principles such as freedom
of expression and association, equality before the law, rights of political
participation—have to be abstract, since they forego appeal to substantive,
culturally specific notions of the good. Yet, if there is to be the mutual trust
required for people not merely to regard such principles as correct, but to
accept the vulnerability, the dependence on the compliance of others and
the subjection to state enforcement, that comes with living in a society
governed by them, they must still understand themselves as indeed “a peo-
ple,” distinct from other peoples and bound together by an antecedent sense
of community independent of their acknowledgment of those principles.4
This common life can no longer revolve around some definition of the
human good. But it can involve such circumstantial factors as geography
and language, and it will certainly and most importantly include a shared
historical experience, centered on the memory of past conflicts, often
bloody, and the hard-won realization that the substantive ideals that once
defined their political existence should now, as abiding objects of reasonable
disagreement, cease to have this authoritative function.
This is why I said at the beginning that liberalism has inherently the
character of a latecomer. It belongs to the very self-understanding of a lib-
eral political order that it has arrived late on the scene. It would be wrong,
of course, to confuse aspiration and reality, to suppose that liberal ideals
have become fully achieved in today’s so-called “liberal democracies” or to
regard “liberal” as simply an historical category designating the practices
Western societies have devised for dealing with reasonable disagreement
about the human good. All the same, liberalism is a political conception
that emerged as a response to one of the defining experiences of modern
times and that has shaped to a considerable extent the world of today. Even
if, as is not true, there were a liberal society without a pre-liberal past, it
4
Liberal thinkers frequently overlook this need for a prior sense of community.
Present-day communitarian thinkers recognize its importance, but wrongly suppose it
can only consist in some shared conception of the good.
66 Charles Larmore
would still define itself by the need to avoid the lure of trying to organize its
political life around some specific religious or ethical vision. Precisely
because liberal principles rule out such ways of setting the terms of political
association, they allude to the temptation of doing so. Liberalism always
presents itself as a better account of the nature of just rule than those we are
or have been initially inclined to adopt.
This does not mean, however, that all the different forms the propensity
to align political society on controversial ideals of the human good can take
have been clear from the start. On the contrary, liberal thinkers themselves
from John Locke’s time to our own have often presented their political phi-
losophy in terms of an overarching individualist ethic, committed to cultivat-
ing a critical attitude toward inherited forms of belief and cultural traditions,
to thinking for oneself and working out on one’s own the life one will lead,
that is far more disputable than they have supposed.
That liberal thought has taken this path is not surprising. Individualism
has formed a powerful current in our world. It grows out of basic features of
modern society, especially the market institutions of a capitalist economy. It
is also an understandable response to the expanding recognition, ever since
the 16th century wars of religion, that reasonable people—by which I mean
people exercising their general capacities of reason in good faith and to the
best of their abilities—tend naturally to disagree about the essence of the
human good. Classical liberals such as Locke, Kant, and Mill found it there-
fore plausible to conclude in their different ways that the principles of polit-
ical society should be based on this individualist ethic, that they should
abstract from controversial ideals of the good in order to express thereby the
spirit that should shape the whole of our lives. Our allegiance, they believed,
to any substantial view of the good—to any concrete way of life involving
a specific structure of purposes, significances, and activities, such as a life
shaped by certain cultural traditions or devoted to a particular religion—
can be truly valuable only if we understand such forms of life as ones we
choose, or would choose, from a position of critical detachment. This gen-
eral individualist perspective formed the framework in which they presented
their political philosophy. (The idea of “individuality” plays this role explic-
itly in Mill’s essay On Liberty.) Our status as political subjects or citizens
should be independent of whatever specific conceptions of the human good
we espouse because in that way political principles mirror—as Locke, Kant,
and Mill might have said respectively—the fallibilist, autonomous, or experi-
mental attitude that we as persons should maintain at the deepest level of
our self-understanding.
Things have not stayed so simple, however. Individualist ideas have
themselves become an object of reasonable disagreement. From parts of
the Romantic movement to present-day communitarianism, there has
developed an appreciation of the significance of tradition to which the

premium that individualism places upon critical reflection appears to
embody a kind of moral blindness.5 Is not a distanced and questioning
attitude toward inherited ways in reality only one value among others? To
give it supreme authority may then block recognition of much else that is
also of value. Thus, it has been held that we can share in the good that
some ways of life offer, only if we do not think of our allegiance to them
as elective, as a matter of decision, but regard it instead as constitutive of
our sense of what is valuable, as rooted in a feeling of belonging. The
importance of common customs, ties of place and language, and religious
faith can lie in shaping the very understanding of good and ill through
which we make the choices we do.
Today, despite the massive influence that individualist modes of thought
continue to wield, they are recognized to be notoriously contentious. No
doubt they have always been contested, but now their difficulties and
drawbacks are widely acknowledged. However we settle to our own satis-
faction the respective merits of thinking things out for ourselves and fol-
lowing tradition, we cannot deny that on this question reasonable people
are bound to disagree. Whether these two outlooks are in fact so sharply
opposed—whether self-reliance is not nurtured by certain kinds of com-
munity and whether any traditional form of life can long endure without
innovation—is but a further dimension of the controversy. Classical liber-
alism aimed at a sort of neutrality—a neutrality of justification, if not a
neutrality of effect—by which the principles of political society would
rest on a basis free from disputable conceptions of the human good.
“Political authority needs only to be just,” Benjamin Constant declared,
“we will take care of our happiness.”6 Yet, the basis adopted was in fact no
less controversial itself.
Here, then, liberalism faces a challenge. Should it stand by its classical
commitment to an individualist view of life, becoming now avowedly a
“perfectionist” doctrine that grounds its political principles on a compre-
hensive ideal of human flourishing? Or should it instead, seeing in this new
area of controversy yet another instance of that tendency to reasonable
disagreement to which it has from the beginning formed a response, seek a
reformulation that can accommodate both sides of the dispute? The second
path is the one pursued by the political liberalism that John Rawls, I, and
5
For a more detailed account of this Romantic theme, see Larmore (1996b),
Chapter 2, as well as Larmore (1996a), 127–34.
6
Constant, De la liberté des Anciens comparée à celle des Modernes (1819): “Que
[l’autorité] se borne à être juste; nous nous chargerons d’être heureux,” in Constant
(1980), 513.
68 Charles Larmore
others have espoused. It seeks to fix the principles of political association in

terms independent, not only of religious convictions and substantive
notions of the good life, but also of ethical ideals defining the sort of
attitude—individualist and self-critical or communitarian and traditional-
ist—that we ought to have toward the conceptions of the good we espouse,
since they too are ineluctably objects of dispute. If the proper aim of gov-
ernment from the liberal standpoint is not to foster some particular idea of
what it is to live our lives well, but instead to create the conditions of jus-
tice—in other words, a fair framework of rules and institutions in which
people can pursue their various ends—then ascent to this more reflective
level of neutrality would appear to be in order. Seen in this light, political
liberalism does not represent a radical departure from the motivations of
its classical forebears. It too aims to define a form of political society that
respects the breadth of reasonable disagreement. The differences stem from
experience, as it has become clearer just how widely reasonable people can
disagree.
Nonetheless, as shown precisely by its commitment to establishing a fair
system of justice, political liberalism would be wrongly understood if one
supposed its paramount objective is simply to secure a basis for political
association about which all reasonable people can agree. Reasonable disa-
greement is the crucial problem to which it offers a political solution. But
the solution itself rests upon moral assumptions that, delineating the basic
conditions of justice and requiring, so it is argued, the setting aside of indi-
vidualist views, play this role because they are held to be right and not
because they are considered to be universally shared. On this essential point,
political liberalism is frequently misunderstood (even by some of its propo-
nents), and one of my principal ambitions is to dispel the confusion. Before
going further in this direction, however, I want to look more closely at the
idea itself of reasonable disagreement. Though it has provided, as I have
said, the main impetus in the development of liberal thought past and pres-
ent, it can—on reflection—seem rather problematic.
2. THE KEY PROBLEM
What, then, is to be understood by reasonable disagreement? It does not

mean, as I emphasized before, the simple fact that people disagree and dis-
agree in particular about the nature of the human good. That is a banality,
familiar from time immemorial. It is, rather, the idea that reasonable peo-
ple, precisely by virtue of exercising their reason in good faith and to the
best of their abilities, tend to come to contrary opinions when they con-
sider, especially in some detail, what it is to live well.
The phenomenon itself seems evident enough. How often, for instance,
do we believe we agree with others about some important matter of life only
to discover, as we discuss the subject further, listen more carefully, and
reflect on what has been said, that we are in reality quite far from being of
the same mind! Since, however, it has long been part of our cultural and
philosophical tradition to believe that reason ensures convergence of opin-
ion, reasonable disagreement can appear to be a notion hard to make sense
of. If people disagree about some subject, given the same evidence, then—
so one can feel impelled to think—they cannot all be reasoning correctly.
How, therefore, can their disagreement count as reasonable? Once they
learn of their disagreement, must they not each, if convinced they are fully
reasonable themselves, reckon that the others are not as reasonable as they?
Or if they believe they are all equally reasonable, must they not each con-
clude that the reasons they supposed they had for their views are inadequate
and that the proper position to adopt, pending new evidence or further
reflection, is to suspend judgment about the matter at hand? At best, rea-
sonable disagreement can thus look like a temporary condition: when peo-
ple who believe they are equally reasonable discover that they disagree, they
would seem obliged to backtrack from their opposing views, and as a result
their disagreement would cease.
A first reply to this objection is that suspending judgment may not be an
option when we are considering, not what to believe about some question
of fact, but which path of life to choose or embrace. Then time and circum-
stance press us to decide one way or the other. We are, moreover, already
embarked on the path that is ours at the moment, and the choice whether
to continue as we have been or to set out in some new direction is one that,
at least implicitly, we cannot escape making.
But in addition and more fundamentally, reasonable disagreement can
be seen as likely and as likely to endure when the subject of discussion, in
whatever domain, is of sufficient complexity and difficulty. When the key
concepts involved can be variously interpreted in plausible but conflicting
ways and when the crucial sorts of considerations that bear on the subject
can be differently weighted, again in plausible but conflicting ways, people
can well come to contrary conclusions if their starting points, that is, the
background of existing beliefs, standards, and interests that appear to them
pertinent, are significantly different. Their disagreement will be reasonable
insofar as they have reasoned properly from their respective points of depar-
ture. Reasonable views are justified views—that is, views one can see reason
to accept, given one’s other convictions—but justified views need not be
true. That is why reasonable disagreement is both reasonable, the different
sides each holding justified positions, and an instance of disagreement, the
different sides presuming that there is a correct answer to the question that
70 Charles Larmore
divides them. Here, by the way, I should point out that “seeing a reason,” as
I use the phrase, entails that there really is such a reason, though a reason
one can grasp from within one’s own perspective, and is thus to be distin-
guished from “having a reason,” which may apply even if one is not in a
position to grasp the reason.
To return to the objection I first mentioned: some may think—as does
much of the current epistemological work on reasonable disagreement7—
that when people who consider one another to be equally reasonable (to be
“epistemic peers,” in the language of that work) realize that they disagree on
some matter, despite having before them the same body of evidence, they
then have good reason to question their opinions, suspend judgment, or
revise their conclusions. For they must suppose that someone is making a
mistake, and if they regard one another as equally reasonable, they each
have no grounds for ascribing the mistake to the others rather than to them-
selves. Reasonable disagreement, again it may be thought, ought to induce
self-skepticism, so that the disagreement itself will end. Yet if, as I have been
arguing, being reasonable consists in believing and doing what is justified
given one’s background beliefs, standards, and interests, then another per-
son’s coming reasonably to a different conclusion than we do is not by itself
a reason for us to think that our view may be false. A good reason for us to
doubt our view must be one we can recognize to be such from our own
perspective, and the fact that another person has come to an opposing view
by reasoning well, but from different starting points than ours, is not in
general that sort of reason. We can regard the other person’s view as reason-
able or justified though at the same tine wrong, because of the mistakes we
see in his underlying premises.
The recent epistemological writings to which I referred can make little
sense of the idea of reasonable disagreement because they slight the way the
most notable examples turn on reasonable people having different back-
ground convictions about how to determine what to believe and do. If they
happen to mention such factors, they often stipulate that people who differ
in this regard cannot regard one another as epistemic peers equipped with
the same evidence; as a result, they define the phenomenon out of existence.
Generally, however, they focus on disagreement in perceptual and mathe-
matical judgments, in which (in contrast to judgments about the human
good, for instance) background convictions play little role.
To this defense of the idea of reasonable disagreement one may rejoin
that, in the situations described, we cannot regard the other person’s view as
justified unless we consider his premises to be justified as well and that we
therefore cannot coherently think of his view as justified though wrong.
7
E.g., Christensen (2007).
Yet, this rejoinder assumes that no proposition can count as justified unless

the beliefs on which it rests are justified as well, and that assumption,
though widespread, is mistaken. Justification, properly understood, is not
an activity in which we engage for its own sake, but rather a response to
some problem, to some question or doubt that has interrupted the course
of our routines. While it is certainly a problem whether to adopt a belief we
do not yet hold or whether to modify or reject an existing belief whose truth
we have discovered reasons to doubt—which is why in these cases justifica-
tion is in order—the mere fact that we possess some belief is not in itself a
problem. We need a reason to open our mind just as we need one to make
our mind up. Beliefs serve to guide our conduct, including the solving of
problems, and the proper object of justification is thus not belief itself, but
changes in belief. Questions of justification arise within a context of given
beliefs that do not in and of themselves need to be justified. Such then are
the terms in which we should judge, not only our own thinking, but the
thinking of others as well. This “contextualist” understanding of justifica-
tion, as I have called it,8 makes the phenomenon of reasonable disagreement
intelligible, and though it is controversial, I repeat the point I made before
(and to which I will return at length): the aim of political liberalism is not
to rise above all controversy.
The factors I have cited to explain reasonable disagreement—that central
concepts can be variously interpreted, pertinent considerations ascribed dif-
ferent weight, especially because the way people reason can be differently
guided by past experience—correspond to the gist of what John Rawls called
“the burdens of judgment” in his own effort to account for reasonable disa-
greement about moral and political matters.9 Rawls acknowledged, however,
that the burdens he described are not peculiar to the practical domain. Indeed,
we may well wonder whether the exercise of reason does not in general breed
disagreement about difficult questions. In pre-modern times, for instance, no
two physicists were likely to think the same who thought at all. If the modern
sciences tend to achieve a convergence of opinion about complicated matters,
might this not be because they have made nature the object of a form of
inquiry, based on controlled experiment and measurement, that is geared to
making them a cumulative enterprise?
What should in any case be clear is that the expectation of reasonable
disagreement in regard to moral questions is not itself a moral doctrine but
instead a conception of reason’s capacities for dealing with these questions.
Sometimes, people (including Rawls himself ) refer to it as “pluralism” or
the “fact of pluralism,” though confusingly so, if an allusion is intended to
8
For more detail, see Larmore (2008), 4–5, 12.
9
Rawls (1996), 56–7.
72 Charles Larmore
the celebrated writings of Isaiah Berlin.10 The pluralism Berlin advocated is

a positive theory about the character of the human good, holding that it
does not derive from one single source of value (such as pleasure, freedom,
or knowledge), but consists instead in a plurality of ultimate ends, irreduc-
ible to any common measure, resistant to any definitive ranking, and liable
to come into conflict with one another. Far from being the same thing as the
expectation of reasonable disagreement, this theory is, not surprisingly, one
more object of such disagreement, a doctrine about whose merits people are
bound to differ.
Perhaps value pluralism, if it is true (as I happen myself to believe), may
help to explain why reasonable people find themselves easily at odds about
what it is to live well: they each, we might surmise, have reason for their
views because they are responding ultimately to incompatible elements of
the good. The explanation of reasonable disagreement in the moral domain
would thus involve, not only some general features of reason, encapsulated
in the so-called “burdens of judgment,” but also the very nature of the good.
That such disagreement should have become so salient and pervasive a fea-
ture of the modern world, however, must depend not on universal, but on
historical factors. Chief among them are surely the great variety of cultural
traditions we have inherited, the increasingly complex forms of division of
labor in which we live, and the modern practices of toleration themselves,
which, in welcoming the expression of differences, have encouraged them
to proliferate further.
Finally, there are two all-important clarifications about the meaning of
the word “reasonable” itself. First, I assume no distinction between “reason”
and “faith.”11 That should be evident from what I said earlier about how
liberal thought took its start chiefly from an appreciation of how readily
matters of religion lend themselves to reasonable disagreement. People of
faith, too, typically seek to understand better their convictions, to interpret
their faith in the light of their experience and to integrate it with their other
commitments. They pursue questions and deal with doubts, and so they
“reason” in the broad sense I intend, even if their distinctive starting points
are allegiance to tradition or belief in revelation.
Second, as my preceding remarks should also have indicated, I mean by
reasonable people those who are exercising their general capacities of reason
10
On the differences between the two, see Larmore (1996a), Chapter 7, and Larmore
(2008), 141–2.
11
In her critical discussion of my views, Martha Nussbaum (2011) complains that I
slight religious citizens by presenting political liberalism as engaged centrally with
disagreement among reasonable people (32). She therefore quite misunderstands me on
this score, presupposing—as in this context I do not—an opposition between reason and
faith. See also n 27.
sincerely and to the best of their abilities.12 Reasonableness, as I use the

term here and in other writings in political philosophy, is thus an essen-
tially epistemic notion. This is not, of course, the only way the term may
be used, or the only way it has been used in talking about liberalism and
its connection to reasonable disagreement. Rawls, in particular, defined
reasonable people (in his book Political Liberalism) as those who both (i)
recognize the burdens of judgment and their consequences for political
association—namely, that only the oppressive use of state power can unify
a society around a single conception of the human good—and also (ii) are
disposed to propose and abide by fair principles of cooperation, given that
others are similarly disposed.13 His notion thus combines an epistemic and
a moral component.
Now, it cannot be denied that in everyday life we sometimes use the term
in the latter, moral sense of showing a concern for fairness (perhaps because
we think of reasonableness more generally as involving an awareness of
human limitations). A person is “reasonable,” we then say, if she not only
can see some disputed matter from the other parties’ point of view, but is
willing to seek some common ground for resolving the conflict. In such
cases, we may even contrast the reasonable person with someone who is
merely “rational,” focused on pursuing his own interests efficiently and
regarding the viewpoints of others as merely further givens of the situation
to incorporate into his calculations. This sort of contrast was highlighted by
W.M. Sibley in an insightful essay that Rawls himself invoked as he went on
to associate our two distinct moral powers, having a sense of justice and
pursuing a conception of our good, with the reasonable and the rational
respectively.14 Because of this difference in what the two of us mean by the
term, “reasonable” as I use it refers to the central problem to which liberal-
ism has sought a political solution—the fact that people reasoning in good
faith and to the best of their abilities disagree about the nature of the
good—whereas Rawls used the term to refer to part of the solution itself.
I have several reasons for not following Rawls’ usage.
If, as I have argued, we cannot fully understand the sources of such disa-
greement in the moral realm without seeing that it is to be expected in any
domain whenever difficult concepts must be interpreted and competing
considerations weighed together, then we need a suitably general notion of
“reasonable” to characterize reasonable disagreement properly. This notion
12
There is a lot more to be said about the nature of reason and of reasons than is
relevant from a political point of view. For my broader and indeed metaphysical views on
this subject, see Larmore (2008), Chapter 5, and Larmore (2012b).
13
Rawls (1996), 48–62.
14
Sibley (1953). Rawls refers to Sibley’s essay in Rawls (1996), 49.
74 Charles Larmore
cannot be specifically moral in character, but must instead be essentially

epistemic.
Furthermore, to let a moralized usage of a not obviously moral term
such as “reasonable” play a fundamental role in one’s political theory—as
Rawls did, for he used it to define many of his key concepts such as reason-
able comprehensive doctrines, reasonable citizens, reasonable pluralism,
and even reasonable disagreement—risks failing to make plain the moral
principles themselves on which the whole theory rests. This is what hap-
pened in Rawls’ case: he never spelled out clearly the moral basis of his
political liberalism.15 In particular, his use of the term “reasonable” to mean
both being disposed to abide by fair terms of cooperation (the moral com-
ponent of his notion) as well as recognizing that, given the burdens of
judgment, a social consensus about the makeup of the good is most likely
due to the pressures of state power (the epistemic component) left obscure
the exact connection between these two components. In virtue of what
moral assumptions does fairness require rejecting that kind of state coer-
cion? I think it best to make as explicit as possible the moral foundations
of one’s political theory. In fact, the assumptions involved are none other
than the basis on which political liberalism seeks to modify the individual-
ist framework of its classical antecedents—namely, a principle of respect
for persons. Having explained the nature of the reasonable disagreement to
which the liberal tradition has all along been a response, I return therefore
to the question of why this phenomenon should now appear to call for
such a revision.
3. FOUNDATIONS
The impetus for the development of political liberalism has been, I observed,
the growing recognition that the general individualist philosophy on which
classical liberalism relied, as it sought to define a fair political order amid
reasonable disagreement about important religious and ethical questions, is
an object of controversy in its own right. However, it is important to see
that political liberalism is heeding more than simply historical experience in
freeing itself from dependence on an outlook toward life that prizes critical
reflection, autonomy, and experimentation. It is also drawing upon moral
assumptions that imply that this is the proper route to take. After all, why
should not liberal thinkers instead dig in their heels and, observing correctly
15
This failure led, I believe, to various deficiencies in his account of “public reason.”
See Larmore (2012a). See also, more generally, my critique of Rawls in Larmore (2008),
Chapter 6.
that no political conception can accommodate every persuasion (a point to

which I shall return), maintain that liberalism stands or falls with a general
commitment to individualism? The answer must be that the core principles
of liberal thought lie at a more fundamental moral level. So the political
liberal needs to make plain what these underlying convictions are.
How far did Rawls take us in this regard? Political liberalism, he once
remarked, “applies the principle of toleration to philosophy itself.” Its goal
is a political order defined by a conception of social justice that is “as far as
possible, independent of the opposing and conflicting philosophical and
religious doctrines that citizens affirm,” including those in terms of which
liberal ideas themselves have often been propounded.16 For only if citizens
can see reason to accept the basis of a political order are they rightly subject
to its authority. This requirement Rawls formulated as the liberal principle of
legitimacy. “Our exercise of political power is proper and hence justifiable,”
that principle holds, “only when it is exercised in accordance with a consti-
tution the essentials of which all citizens may be reasonably expected to
endorse in the light of principles and ideals acceptable to them as reasonable
and rational.”17
This is, of course, a consent principle of a sort not unfamiliar in the lib-
eral tradition. We should note, however, that it makes the legitimacy of
political rule depend not on the actual but on the hypothetical consent
of the governed and doubly so, referring to political arrangements that peo-
ple can (not necessarily do) see reason to accept and can do so insofar as
they are (in Rawls’ sense of the terms) reasonable and rational. On the latter
point there will be much more to say in the next section. Rawls’ principle
also, I agree, gives expression to the abiding heart of the liberal vision, the
core commitment in virtue of which its long-standing reliance on individ-
ualist forms of thought needs to be abandoned. Yet, it does not take us
deep enough. For what are the reasons that favor this understanding of
political legitimacy in preference to others which, historically, have appealed
instead to the will of God, the ways of the forefathers, the cultivation of
human excellence, or indeed the development of individuality? We need to
bring out, as Rawls himself did not do, the moral foundations of this idea
of legitimacy.
Rawls (1996), 9–10.

16
Rawls (1996), 217; see also 137. Here, it is to be noted that “legitimacy” and
17
“justice” are not synonymous, as shown by the fact that laws may be legitimate without
being just and just without being legitimate. The difference between them is that
legitimacy applies to a political regime as a whole and has to do with its exercise of
coercive power being justified—though the two terms are not entirely disjoint since the
conditions under which coercive power may rightly be exercised form part of the domain
of social justice. For more on this point, see Larmore (2013), 293–4.
76 Charles Larmore
These assumptions, as I suggested at the end of the previous section,

provide the link, never explained by Rawls, between the two components,
moral and epistemic, of his notion of the reasonable. If reasonable people
are to be characterized as combining a commitment to fair terms of social
cooperation with the recognition that only the oppressive power of the
state, not the free use of reason, can produce a social consensus about some
comprehensive religious or ethical doctrine (including individualist views of
life), it must be because they would believe that working out fair terms of
cooperation involves finding rules of political association that people can by
their own lights see reason to accept. This is what Rawls himself says: “rea-
sonable persons . . . desire for its own sake a social world in which they, as
free and equal, can cooperate with others on terms all can accept.”18 In other
words, being reasonable (in Rawls’ sense) supposedly entails adhering to the
“liberal principle of legitimacy.” Yet, why should that be so? Why should
fairness require consent? We cannot answer that question unless, once again,
we make clear the moral basis or substance of that principle itself. Only
then will it become clear why political liberalism regards it as imperative to
go beyond its classical forebears.
In earlier writings I have argued that this moral basis lies in a certain idea
of respect for persons.19 It is crucial, however, to perceive what idea of respect
is involved, since there are many different concepts in circulation, some of
them entailing comprehensive ethical doctrines of the sort that political
liberalism must want to avoid. That is one reason to go through the argu-
ment again, though I will this time add more detail as well.
To begin, let us ask why we believe (supposing we do) that the funda-
mental terms of political society should be such that those whom they are
to bind, understanding themselves, as Rawls says, to be free and equal—
that is, to be each endowed with the two moral powers (of developing a
sense of justice and a conception of the good) and to be each members of
the community in their own right, not antecedently subject to the authority
of another—can see reason to agree to them. Do we accept this principle
because we think that in general people are bound only by moral rules they
could not reasonably reject?20 There are a number of difficulties with this
response. One is that the conception of morality to which it appeals looks
circular. Can there be good reasons to accept the basic rules of morality that
do not themselves turn on the fact that it is simply wrong not to observe
them? Is it not their rightness which explains why they are not to be reason-
ably rejected, and not the reverse? Or if that is not so, then does not such an
18
Rawls (1996), 50.
19
See, in particular, Larmore (2008), Chapter 6.
20
Such a theory of morality is presented in Scanlon (1998).
account of morality show an intimate connection with those individualist

ideals of self- determination to which, for political liberalism, the ground
rules of political society need to be neutral? Should we really not judge
others by moral rules that they from their perspective would find unaccept-
able? In any case, whatever the strength of these objections, views about the
basis of moral rules in general are not, I believe, the pertinent source of our
conviction that political principles must be able to meet with the reasonable
agreement of the citizens they are to govern.
That conviction reflects instead the distinctive feature of political prin-
ciples themselves that sets them apart from the other moral rules to which
we hold people accountable, namely that they are coercive. Moral rules may
be divided into two groups. With some we believe people can be rightly
forced to comply, whereas others we do not regard as valid objects of legal
enforcement, whatever disapproval or even outrage we may feel when they
are violated. The first group alone has the status of political principles. For
an association is political insofar as it possesses (or asserts it possesses) the
means to secure compliance with its rules by way of what it claims to be the
legitimate use of coercion.21 Observe that the use or threat of force cannot
be deemed inherently wrong, for then political association would be impos-
sible. From a political perspective, the question is rather the conditions
under which coercion is justified or legitimate. Different political systems
invoke different sorts of justification—different legitimation stories—for
their authority to impose, by force if necessary, the rules defining their
terms of association. The characteristic claim of a liberal order is, roughly
speaking, that its legitimacy in imposing coercive rules of political associa-
tion lies in its citizens being able to see reason to accept these rules. Rawls’
“liberal principle of legitimacy,” for instance, refers to “the exercise of polit-
ical power” as what is to be governed by the reasonable agreement of citi-
zens. Given then that the legitimate use of coercion is the problem, what
are the moral assumptions that justify the solution that liberalism charac-
teristically urges?
They have to do, I believe, with an essential feature of what it is to be
a person. As persons we are, whatever our view of the good, beings essentially
capable not only (as are the higher animals too) of thinking and acting for
what we take to be reasons, but also of reflecting on such reasons in the sense
of examining whether what appear to be reasons really are good reasons.
Reflection of this sort, let it be noted, is an activity in which we can engage
from a variety of standpoints and not just in an attitude of critical detach-
ment in which we stand back from inherited forms of life in order to work
21
Here I follow Max Weber (1972), I.1.§17. For more on this conception of the
political, see Larmore (2013).
78 Charles Larmore
out for ourselves how we should think and act. People of faith, people
imbued with a sense of tradition, still reflect on the real worth of the reasons
their commitments appear to give them—they often cannot help but do
so—even if their deliberations take place within the bounds of their guiding
assumptions. Nothing in this concept of a person or in the principle of
respect that, as we shall see, builds upon it and grounds the liberal idea of
political legitimacy harbors an allegiance to individualist ideals.
To perceive the connection between these three concepts—person, respect,
and liberal legitimacy—imagine a situation in which we seek to bring about in
others by the threat of force their conformity to a rule of conduct we do not
suppose that they too could see reason from their point of view to impose on
the community. We would, it is true, be appealing to their ability to think and
act for what they take to be reasons, for they cannot be moved by threats
except by seeing reason to fear what we may do if they fail to comply. But our
aim would be only to take advantage of that ability, compelling them thereby
to act in a way that advances the ulterior goals we seek to achieve: the estab-
lishment of social order, the realization of some public good, even the satisfac-
tion of some interest of theirs they cannot properly pursue themselves. We
would not then be valuing their ability to think and act for reasons as some-
thing good in itself. For that would involve treating them in such a way that
they could be moved to abide by the rule, not by the reasons to fear the con-
sequences of non-compliance, but by the very reasons we ourselves have for
imposing the rule. In other words, we would have to appeal to their distinctive
capacity as persons, namely their ability to reflectively evaluate apparent or
purported reasons. If we thus sought to engage their capacity for reflection in
getting them to adhere to some rule of conduct, we would be showing this
capacity of theirs precisely the same intrinsic regard we show our own when we
propose to heed that rule in virtue of concluding that there are good reasons
to enact it. As one might say in echo of Kant, though without the metaphysi-
cal complexities of Kant’s ethics and also (unlike him) with an eye only to
coercive actions, we would then be treating these persons, in their capacity as
rational and reflective beings, no longer solely as means but also as ends.
To respect others as persons in their own right when coercion is involved
is therefore to require that political principles, or at least the fundamental
ones, be as justifiable to them from their perspective as they presumably are
to us.22 We need not suppose the same to hold for those moral principles by
which we evaluate others, even strongly disapprove of how they may act, yet
do so without appealing to the state’s means of coercion; those principles,
we may believe, are morally binding on them, whether or not they can see
22
In the next section I explain the qualification as well as indicate how this initial
formulation itself needs to be modified somewhat.
reason to accept them. Why, indeed, should a commitment to liberalism

inhibit our powers of moral judgment? But the use or threat of force is dif-
ferent, since it consists in compelling people to do what they otherwise
would not do and it would, if unconstrained by the norm of respect, involve
treating them, in their capacity as rational and reflective beings, merely in
an instrumental way. To be legitimate, so the principle of respect asserts, the
use or threat of force must accord with the reason of the citizens themselves
who are potentially subject to it.
Respect for persons in this sense forms, then, the underlying moral basis
of the liberal principle that the fundamental rules of political association
are legitimate only if they can be the object of reasonable agreement
among those whom they are to bind. To this account, which bases respect
on the capacity for reason, it may be objected that a liberal regime should
treat people with respect even when they happen to be unreasonable,
even when they are carried away by passion or pursue their ends in counter-
productive ways. That is certainly true. But how else should such a regime
show them respect except by holding them to be subject, even then, only to
those coercive rules they could, when thinking more clearly from their
point of view, see reason to accept? Thus we see plainly why the defining
problem for liberalism is reasonable disagreement, not mere disagreement,
about the human good. The problem of what reasonable people, despite
their inevitable disagreements, can nonetheless agree to as the basis of polit-
ical authority sets the very terms of the liberal ideal of just rule. Here as
elsewhere we can understand the nature of liberalism only if we keep clearly
in focus the problem to which it aims to be the solution.23
The concept of respect for persons I have outlined is not, to be sure, the
only sense the notion of respect can have. Nor is it all that we might under-
stand by respect in a comprehensive moral theory. But it is the principle
that lies at the foundations of a liberal conception of political society.24 It is,
23
In arguing that reasonable disagreement is no more of a political problem than
simple disagreement, Joseph Raz (1998) misses this connection between problem and
solution.
24
Rawls himself showed a great reluctance to admit that a principle of respect lies at
the basis of liberal thought, largely because of the many different meanings the idea of
“respect” can have. At the end of Rawls (1971), for instance, he declared that he had not
sought to derive his principles of justice from the notion of respect for persons because
that very notion calls for interpretation, which only a conception of justice can provide
(585–6). The hermeneutic point is well taken. But it does not rule out the possibility that
respect, in a sense to be grasped only in the light of his theory as a whole, is a value on
which that theory substantively depends. We may have to rely on our thinking about
justice or about political legitimacy in order to determine the appropriate sense of
“respect” they embody. Yet, the principle of respect for persons, thus delimited, may still
form the moral foundation of the doctrines themselves.
80 Charles Larmore
in particular, the principle on which political liberalism relies in arguing

that the individualist assumptions of its predecessors should be abandoned.
For not only does this idea of respect, as I indicated, carry no allegiance to
individualist views of life; it also requires, given the persistence of reasonable
disagreement about such views, that they play no role in shaping the basic
rules of political society.
Since then political liberalism builds upon this moral foundation, we
would be wrong to think it accords respect for persons (in regard to coercive
principles) the political significance it does because citizens can all see rea-
son for doing so. That supposition is not in fact likely to be true. Reasonable
people in my sense of the term—exercising their general capacities of reason
in good faith and to the best of their abilities—need not, depending on
their beliefs and interests, regard it as imperative that coercive principles be
rationally acceptable to those whom they bind. They may on the contrary
consider themselves justified in rejecting that notion if they believe, for
instance, that the most important feature of political society is that it be
pleasing to God. Of course, if “reasonable” is understood (more narrowly)
in Rawls’ sense, then such people will see good reason to believe that polit-
ical principles must respect those whom they are to bind. But that, as I have
explained, is trivial, since true by definition.
No, respect for persons has the position in political liberalism it does, not
because it constitutes common ground and forms an object of reasonable
agreement, but because it is what directs us in the first place to look for
common ground, to seek the principles of our political life in the area of
reasonable agreement. Thus, political liberalism does not seek to correct the
individualism of its classical antecedents simply because individualist views
of human flourishing have shown themselves to be eminently controversial.
Their disputability is seen as calling for such a revision only in virtue of a
commitment to the principle of respect.
4. IMPLICATIONS AND PROSPECTS
This foundational role of the principle of respect has a number of important

implications.
One is, as I just intimated, that it is not the ambition of political liber-
alism, contrary to what many suppose, to show that all citizens or people
generally can in the light of their deepest interests and convictions see reason
to endorse its defining principles. It is naïve to suppose that any political
ideal, given the diversity of views about the good and the right, can be in
this sense universally justified. Such is also the position of Jonathan Quong
in rejecting what he calls an “external” conception of political liberalism.
Yet, I am unhappy with the way he distinguishes the “internal” conception

he favors, as when he asserts:
The aim is not to justify liberalism to a radically diverse constituency, one that may
include people who reject liberalism’s most fundamental values. Rather, the aim is
to understand how liberal rights and institutions can be publicly justified to the
constituency of an ideal democratic society.25
Though political liberalism must certainly explain how citizens who hold
liberal values can do so compatibly with disagreeing about so much else, it
is crucial to remember—I repeat a point I have made many times—that
liberalism in general and political liberalism in particular have taken shape
as a response to a problem, namely the political problem posed by “radical
diversity” or by what I have been calling reasonable disagreement.26 Quong
tends to proceed as though the liberal tradition already exists, the question
being whether its guiding ideals and assumptions can be shown to be coher-
ent. But the deeper question is why it exists at all, what problem it has arisen
in order to solve. Without an answer to that question, we cannot under-
stand its underlying motivations, which are what I have been trying to bring
out.27 Still, it is true—on this Quong is right: the solution it offers is not one
that can necessarily be justified to all reasonable people, understood as peo-
ple exercising their rational faculties about moral questions sincerely and to
the best of their abilities. In this, it is unexceptionable since every political
conception excludes, a point to which I return later on.
A second implication is that in a liberal regime the norm of respect does
not have the same sort of standing as the constitutional principles worked
out on its basis. Those principles are legitimate in virtue of being able to be
the object of reasonable agreement and their political authority can accord-
ingly be said to derive from the collective will of the citizens. But the same
is not true of the idea of respect. It must be understood as having a deeper
kind of validity. To be sure, it constitutes a political principle since its
25
Quong (2011), 6.
26
Quong goes so far as to assert that on the internal conception “pluralism [he means
reasonable disagreement, CL] is not a fact about the world which liberal theory must
accommodate. Rather, pluralism is to be understood to be a consequence of liberalism
itself ” (Quong (2011), 139; also 142). Though it is true, as I mentioned earlier, that
liberal ideas of toleration have helped to expand the breadth of reasonable disagreement,
they originally emerged as a response to reasonable disagreement about the nature of the
human good.
27
This is why I also think it a mistake for Martha Nussbaum (2011: 20) to want to
discard from the liberal idea of respect its epistemic component, the reference to what
Rawls called “the burdens of judgment” or I the prevalence of “reasonable disagreement.”
One thereby loses a sense of the problem to which that idea is a response.
82 Charles Larmore
requirement that the terms of political life be reasonably acceptable to all is

itself a rule to be enforced. But unlike the other political principles it serves
to ground, it does not draw its authority from citizens being able to see a
reason to endorse it. Respect for persons must be regarded as a principle
binding on citizens independent of their collective will, enjoying an author-
ity they have not fashioned themselves. For only in recognizing its authority
are they moved to give their political existence the consensual shape it is
meant to have. At the fundamental level, citizens of a liberal democracy
cannot, contrary to Jürgen Habermas, regard themselves as the authors of
all the laws to which they are subject.28
What, then, is the source of the authority of the principle of respect for
persons? What, we may more specifically ask, is its justification absolutely
speaking, the underlying basis on which it rests, as opposed to the different
sorts of reasons citizens may see in the light of their own beliefs and interests
to accept it? I do not have an answer to this second question. But I caution
that we should not suppose that there must be one. At some point, justifi-
cation in this sense comes to an end and we must recognize that certain
principles ultimately speak for themselves. My aim here, in any case, is to
point out the foundational role that respect plays in political liberalism and
the consequences that follow from this fact.
Now a third, very important implication is that the idea of respect serves
to define the very nature of the consensus that counts as justifying the impo-
sition of political principles. To see this, observe first that agreement in this
case is manifestly a normative notion, if only because it means reasonable,
not simply actual agreement. Political life, from the liberal standpoint, is to
be based on principles which citizens, despite their various moral, religious,
and metaphysical beliefs, can from their perspective see reason to accept.
Though what they can see reason to accept depends on their existing beliefs
and interests, these reasons must (for the system to be legitimate) be reasons
they really have, not reasons they merely think they have. However, reason-
ableness in this sense is not the sole standard or constraint governing the
consensus in question. Because the idea of respect for persons is what
requires the search for common ground, it must figure as a further condi-
tion that political principles have to satisfy if they are to count as being the
object of reasonable agreement among citizens. In other words, the terms of
political society are to be judged by reference to what citizens would accept,
were they not only reasonable but also committed to the norm of equal
respect for persons.
This means that the moral idea of respect for persons lying at the heart
of political liberalism should be formulated more precisely as follows: the
28
Habermas (1992), 51f, 135, 153.
fundamental principles of political society, being coercive in nature, ought

to be such that all who are to be subject to them must be able from their
perspective to see reason to endorse them on the assumption—perhaps coun-
terfactual (I will come back to the import of this clause)—that they are
committed to basing political association on principles that can meet with
the reasonable agreement of citizens. The constraint that the idea of respect
places on reasonable agreement as the basis of legitimate political principles
is implicit in Rawls’ own “liberal principle of legitimacy.” For recall that,
according to that principle, the exercise of political power is justifiable only if
it rests on principles acceptable to all citizens insofar as they are, Rawls stipu-
lated, “rational and reasonable,” and what he meant by “reasonable,” as we
have seen, is being disposed to seek fair terms of cooperation, that is, terms
justifiable to all. Yet, the moral content of the constraint lies hidden in Rawls’
formulations. I believe clarity is served by bringing it out explicitly.
These three implications of the foundational role of the principle of
respect help us to see, fourthly, what is wrong with a frequent objection to
political liberalism. Ideas of social justice, it is objected, are no less subject
to reasonable disagreement than ideas of the good and therefore the search
for principles of political association to which all citizens can reasonably
agree must come to naught.29 It is, however, a misunderstanding of political
liberalism to suppose that its aim is to develop a framework of justice that is
uncontroversial. In general, we should note, conceptions of justice have a
different function than conceptions of the good: they define the framework
of rules and institutions in which people may pursue their various and
sometimes conflicting ideas of what gives life meaning. Such a framework is
just, according to political liberalism, if it rests upon the idea of respect for
persons, and that idea requires, as we have now seen, that the fundamental
principles of political society be ones that citizens can see reason to endorse
on the assumption that they accept this very idea. At the same time, it is
acknowledged that some citizens may see from their perspective no reason
to accept it. Political liberalism, properly understood, is under no illusions
on this score. True, such people, who stand opposed to the very notion of a
liberal society, will still be obliged, by force if necessary, to comply with its
rules even though they may see no reason to endorse them. But this, as my
reformulation of the idea of respect suggests, is not so much to fail to respect
them as it is to compel them to act in accord with the rules of a political
community that they would endorse if they themselves accepted the idea
of respect.
There is another source of continual disagreement about what is just. It
is the fact that the notion of respect itself does not determine one single
See, for instance, Waldron (1999), 105f, 112, and Chapter 7.

29
84 Charles Larmore
conception of justice. It delimits a range of eligible conceptions, what Rawls

called “the family of liberal conceptions.”30 So, it is to be expected that a
liberal society will be home to ongoing debate, within such bounds, about
the specific character of social justice.31 On some basic rights and freedoms
the different conceptions will agree, and insofar as these matters appear
entailed by the very idea of respect, they should be the object of constitu-
tional guarantees. Yet, political decisions must be made that go beyond
these points of agreement, and here the different liberal conceptions are
bound to come into conflict. That is why I said earlier that the idea of
respect properly applies only to the fundamental principles of political soci-
ety. But it is also why one of those principles must be that such decisions
of policy (as we may call them) are to be made on the basis of democratic
procedures (e.g., elections, majority voting) in which all citizens have an
equal say. In this way, citizens who disagree with the rules thus enacted, but
with which they will still have to comply, are shown respect at least indi-
rectly in virtue of their seeing reason to accept these procedures. This is one
reason why political liberalism entails the necessity of democratic self-rule.
Yet, a fifth crucial implication of the foundational role of the idea of
respect concerns a point which I have already noted, but which is worth
exploring further—the fact that a commitment to respect for persons may
not form part of every citizen’s thinking or be a commitment every citizen,
given his present thinking, can see reason to adopt. The idea that the coer-
cive principles of political association should be justifiable to those whom
they are to bind is perhaps so widespread in modern liberal societies as often
to go without saying. But it was not so in the past. And in today’s world it
is rejected, perhaps most notoriously, by those (of whom there are many not
only in parts of the world other than the West) who believe that in the
political realm conformity to God’s will overrides respect for human rea-
son.32 It is also rejected by those who, wedded to certain ethical ideals,
believe a state that fails to devote itself to fostering virtue and excellence, as
they conceive them, is unworthy of allegiance. In virtue of the natural ten-
dency to reasonable disagreement about moral matters, we should not be
surprised that so many in good faith refuse to accept the liberal vision of
political life.
30
Rawls (1996), xlviii, 6, 223.
31
It is also to be expected that different liberal societies, as they come to make respect
for persons their fundamental commitment, will develop as a result of their specific
historical experience somewhat different interpretations of what it entails for the
constitution of a just society.
32
This perspective is to be distinguished from the religious view, very influential in the
historical development of liberal thought, that God’s will is that political society take
shape in accord with human reason.
Thus, we need to recognize candidly that the inclusiveness to which lib-

eral societies aspire also excludes. It denies that the basic terms of our polit-
ical life must be justifiable to citizens who reject the cardinal importance of
the search for common ground amid different convictions about the essence
of the human good. That modern liberalism is a system of exclusion is a fact
that liberal thinkers and citizens alike have often had difficulty acknowledg-
ing. From its beginnings in the early modern struggle for religious toleration
to the present day, the liberal tradition has sought to build a truly inclusive
community. It tends to forget that every principle of political inclusion,
however broad-minded it may be, constitutes a principle of exclusion as well.
For every such principle embodies values that some people reject and con-
sequently defines a community from which they are shut out or in which
they must feel like strangers. This is the important truth that Carl Schmitt
discerned, among so much else he got wrong, when he declared that politi-
cal society depends essentially on a distinction between friend and enemy.33
No political conception, not even political liberalism for all its desire for
accommodation, can prove agreeable to every view of the human good or of
what gives life meaning.
The liberal ambition cannot, therefore, be, as I observed at the beginning
of this section, to establish a political order that does not exclude, that is
acceptable to all reasonable people. It cannot be to bring about a general
reconciliation of individual liberty and political authority, for that is impos-
sible.34 The goal must be to institute a political order that practices the right
kind of exclusion—which means, from the liberal standpoint, one that rests
on the principle of respect for persons, holding itself accountable therefore
only to those who are committed to regulating the political use of coercion
by that very principle.
At the same time, it is important to realize that those whom political
liberalism excludes it does not exclude altogether. It is not the entirety of
their beliefs and interests to which it refuses to grant recognition or allow
any weight in defining the shape of political society, but only that part of
their outlook which entails a rejection of the liberal idea of respect for per-
sons. Honoring the spirit of that idea, seeking as far as possible to keep the
use of coercion within the bounds of what those who are subject to it can
Schmitt (1932).
33
Rousseau held that devising a form of association that reconciles political authority
34
with the liberty of each citizen is the “fundamental problem” of politics (Du contrat social,
I.6). This problem is insoluble, contrary to what many liberal thinkers—such as recently
Gerald Gaus (2011: 1–2)—have supposed. The liberal aim is instead to reconcile political
authority with the liberty of those committed to certain fundamental moral principles.
Rousseau himself may have been thinking along such lines, as suggested by his contrast
between natural and moral liberty at the end of Chapter I.8 in the same work.
86 Charles Larmore
see reason to accept, requires (as indicated in my final statement of the idea)
ensuring that the basic principles of political life be justifiable to such peo-
ple as well, on the assumption—counterfactual in their case—that they too
held this sort of respect to be a fundamental commitment, but given
everything else in their present perspective that they could, compatibly with
that, continue to affirm. Though political liberalism, like every political
conception, inevitably excludes, it also includes the excluded in a qualified
way, and not merely as outsiders to be abided (as other conceptions may
do), but as citizens with a role to play in shaping the terms of association.
In this regard, which is not the least of its appeals, it is, I think, unique
among forms of political exclusion.
I cannot, however, end on so positive a note. There remains a worry,
connected with the argument with which I began this essay, about whether
political liberalism is able to meet the challenges of the contemporary world.
Liberalism in general, I observed, is by its nature and in its self-understand-
ing a latecomer in the history of political conceptions. In explaining that
point, I signaled an important assumption in the liberal project of detach-
ing the rules of political society from substantive religious and ethical prem-
ises and recasting them as a body of abstract principles, first and foremost
that of equal respect for persons as well as those of freedom, equality, and
democratic self-government that derive from it. For that project to succeed,
I argued, the citizens of a liberal society must still regard themselves as form-
ing one people, bound together by a solidarity that is rooted now primarily
in their historical experience, if they are to be willing to expose themselves
to the vulnerabilities of political association. Not by accident, liberalism
developed as a political conception keyed to the modern nation-state.
The question, however, is whether this underlying assumption is still real-
istic and whether the liberal tradition, including political liberalism, has the
resources to deal with the salient problems of our time. In a world dominated
by a rampant globalized capitalism and hurtling, not unconnectedly, toward
environmental disaster, the idea of distinct peoples and sovereign nation-
states appears to be increasingly a thing of the past. States find themselves
more and more the prisoner of worldwide financial markets and multina-
tional corporations. The new digital technology of the media and the Internet,
an integral part of the dynamics of international capital, acts to homogenize
cultural differences and to erase historical memory with the immediacy of
information and entertainment on command. Meanwhile, the air thickens,
the ice-packs melt, the seas rise, and existing nation-states, such as they are,
seem an irrelevance, if not indeed an obstacle, to any way of averting the
impending catastrophe.
Some have thought that the solution to these global problems lies in
globalizing the liberal perspective itself, in working out theories of global
justice. I am not convinced. Political principles of an abstract liberal charac-

ter, whatever may be the scale of their deployment, are viable only if the
people to be bound by them already share feelings of mutual trust and
regard themselves as having a common destiny. It is not clear that humanity
can ever form a people in this sense. Certainly, individuals have not in the
past been protected simply by respect for their “human rights,” independently
of particular states finding it in their interest to intervene. I suspect that
similarly conceptions of global justice, whatever their moral merits, have a
chance of being implemented only if states, liberal states, find themselves
moved to put them into practice. Yet, how likely is that in the present age?
The distinctive problems of our world are not among the problems for
which liberalism was devised, and they threaten its very viability. Its pros-
pects, I am sad to say, are accordingly uncertain.
Bibliography
Christensen, David (2007). “Epistemology of Disagreement: The Good News,”
Philosophical Review 116, 187–217.
Constant, Benjamin (1980). De la liberté des modernes (Paris: Pluriel).
Gaus, Gerald (2011). The Order of Public Reason (Cambridge: Cambridge University
Press).
Habermas, Jürgen (1992). Faktizität und Geltung (Frankfurt: Suhrkamp).
Larmore, Charles (1987). Patterns of Moral Complexity (Cambridge: Cambridge
University Press).
Larmore, Charles (1996a). The Morals of Modernity (Cambridge: Cambridge
University Press).
Larmore, Charles (1996b). The Romantic Legacy (New York: Columbia University
Press).
Larmore, Charles (2008). The Autonomy of Morality (Cambridge: Cambridge
University Press).
Larmore, Charles (2012a). “Political Liberalism and Public Reason: A Critique of
John Rawls,” in Die Welt der Gründe. XXII. Deutscher Kongress für Philosophie.
Deutsches Jahrbuch für Philosophie, Band IV (Hamburg: Felix Meiner Verlag),
1249–62.
Larmore, Charles (2012b). Vernunft und Subjektivität (Berlin: Suhrkamp).
Larmore, Charles (2013). “What is Political Philosophy?,” Journal of Moral
Philosophy 10, 276–306.
Nussbaum, Martha (2011). “Perfectionist Liberalism and Political Liberalism,”
Philosophy & Public Affairs 39, 3–45.
Quong, Jonathan (2011). Liberalism without Perfection (Oxford: Oxford University
Press).
Rawls, John (1971). A Theory of Justice (Cambridge, MA: Harvard University Press).
Rawls, John (1996). Political Liberalism (New York: Columbia University Press).
88 Charles Larmore
Raz, Joseph (1998). “Disagreement in Politics,” American Journal of Jurisprudence

43, 25–51.
Scanlon, T.M. (1998). What We Owe to Each Other (Cambridge, MA: Harvard
University Press).
Schmitt, Carl (1932). Der Begriff des Politischen (Munich: Duncker & Humblot).
Sibley, W.M. (1953). “The Rational versus the Reasonable,” Philosophical Review
62, 554–60.
Waldron, Jeremy (1999). Law and Disagreement (Oxford: Oxford University Press).
Weber, Max (1972). Wirtschaft und Gesellschaft (Tübingen: Mohr).
4
Political Liberalism, Political
Independence, and Moral Authority
Dale Dorsey
As I use the term here, political liberalism holds that an acceptable theory
of political justice cannot be guided by substantive principles of the good
about which reasonable people may disagree. If a particular policy or
institutional action is a requirement of justice, this cannot be because this
policy or action promotes human welfare, well-being, or flourishing.
Morality, on the other hand, isn’t like this. Morality doesn’t bracket sub-
stantive considerations about the value of human life in assigning moral
obligations. It seems perfectly sensible to say that if some action of mine
would render many people’s lives more valuable, this is certainly a moral
reason to do it.
These two facts form the central tension I explore in this chapter. Political
liberalism, or so I claim, comes out with the short end. The general argu-
mentative structure runs like this: first, under the sort of political liberalism
at issue here, individuals can be subject to conflicting political and moral
obligations. At a given time, someone can be morally required to φ and
required as a matter of political justice not to φ. Second, I argue that this
fact creates a problem for political liberalism. Liberals are forced to either
reject a sensible claim about the rational authority of moral requirements,
or adopt a position that is incompatible with one of liberalism’s central
motivations.
1. POLITICAL LIBERALISM AND POLITICAL REASONS
Before I begin the substantive argument of this chapter, I want to character-

ize in a little more detail the view I call “political liberalism.” Political liber-
als accept what Steven Wall calls “the bracketing strategy”:
90 Dale Dorsey
[I]n modern western societies there exists a multitude of conflicting, irreconcilable

and reasonable religious, moral and philosophical doctrines. Taking this “fact” as a
permanent feature of these societies, the [bracketing strategy] seeks to establish that
it is unreasonable or inappropriate to use political power to advance controversial
conceptions of the good society or to promote controversial ideals or values.
Naturally, this claim leads proponents of the [bracketing strategy] to search for a
conception of politics that all parties could reasonably accept, despite the beliefs and
ideals that divide them. (Wall 1998, 29.)1
The bracketing strategy holds that theories of political justice must be neu-
tral with respect to substantive moral, religious, or evaluative principles. Of
course, different versions of political liberalism will adopt very different
forms of the bracketing strategy. Some will hold that only facts about welfare
or the good life ought to be bracketed. Others hold that not only should facts
about welfare be bracketed, but also any or all substantive action-guiding
(such as moral) considerations. For the purposes of this chapter, I construe
the bracketing strategy as narrowly tailored: I discuss only those views that
bracket facts about the value of human life, or about what makes for a good
life rather than a bad one. Insofar as this covers most all versions of political
liberalism, however, I take it that this is not much of a limitation.
One can clearly see the reasoning for a bracketed theory of political jus-
tice in Rawls. Broadly speaking, Rawls holds that we have reasons to value
theories of justice that are stable. In A Theory of Justice, Rawls introduces the
idea of stability this way:
Now a well-ordered society is also regulated by its public conception of justice. This
fact implies that its members have a strong and normally effective desire to act as the
principles of justice require. Since a well-ordered society endures over time, its
conception of justice is presumably stable: that is, when institutions are just (as
defined by this conception), those taking part in these arrangements acquire the
corresponding sense of justice and desire to do their part in maintaining them. One
conception of justice is more stable than another if the sense of justice that it tends
to generate is stronger and more likely to override disruptive inclinations and if the
institutions it allows foster weaker impulses and temptations to act unjustly.
(Rawls 1971, 454.)
For a conception of justice to meet the test of stability, it has to be that a
society well ordered by its principles yields citizens who act on political
requirements not simply for pragmatic reasons or as a simple modus vivendi
(Rawls 1995, xxxix–xl). It must be that individuals are not just motivated to
act according to the principles of justice, but that each “supports the politi-
cal conception for its own sake, or on its own merits,” (Rawls 1995, 148).
The latter requirement is important: for a conception of political justice to
1
Michael Sandel also uses the “bracketing” terminology in Sandel (1989).
Liberalism, Independence, and Moral Authority 91
be genuinely stable, it cannot be the case that it would be abandoned if

factual circumstances change—which would be precisely the case if citi-
zens value this conception of justice strictly as a compromise, or as a modus
vivendi, or as strictly instrumental to the achievement of other things they
care about.
However, this test will be failed by any non-bracketed political theory of
justice. Rawls’ central reasoning for this claim is the “fact of reasonable
pluralism”: there exists a broad array of reasonable comprehensive doctrines.
Given this, agreement on any particular substantive view of, e.g., human
flourishing will not be possible in modern constitutional democracies.2
Thus, the importance of stability, coupled with the fact of reasonable plural-
ism, yields the result “that the state is not to do anything intended to favor
or promote any particular comprehensive doctrine rather than another, or
to give greater assistance to those who pursue it . . . [B]asic institutions and
public policy are not to be designed to favor any particular comprehensive
doctrine,” (Rawls 1995, 193–4). Thus, for Rawls, no stable theory of politi-
cal justice will allow institutions to promote controversial views of the
nature of human flourishing or well-being (Rawls 1995, 133–4). The argu-
ment from stability to the bracketing strategy is put here succinctly by
Jonathan Quong: “Political liberals claim that reasonable citizens will inev-
itably hold different conceptions of the good life or human flourishing. If
liberalism is to be stable for the right reasons . . . the congruence of justice
and the good cannot be understood in terms of any single account of human
flourishing,” (Quong 2011, 164).
Of course, it’s worth noting that Rawls states that stability is not a decisive
reason to choose a particular theory of justice rather than another. Stability
is a theoretical virtue, but it’s not the only one. (See Rawls 1971, 453–7.)
However, for present purposes this point is neither here nor there. It’s clear
that while stability for the right reasons may be only one among potentially
many virtues of a theory of justice, it is a very important argument in favor
of the bracketing strategy. And, hence, any bracketed theory of political jus-
tice should seek to accommodate this constraint.3
The precise contours of political liberalism are the subject of some con-
troversy and confusion. (See, most importantly, Sher (1997), Ch. 2.) But the
view I address here, which is reflected in the ideas just canvassed, can be stated
somewhat more precisely as follows. Take the idea of a reason. Reasons are
the sort of thing that count in favor of particular actions: they are consider-
ations that help to determine deontic valence. Of course, reasons come in
2
According to Rawls, the fact of reasonable pluralism is “the inevitable outcome of
free human reason.” Rawls, Political Liberalism, 37.
3
Thanks to an anonymous reviewer for highlighting this point.
92 Dale Dorsey
many different flavors. There are moral reasons, prudential reasons (reasons,
i.e., of self-interest), aesthetic reasons, reasons of etiquette and protocol, etc.
In each of these individual domains, the relevant reasons will help to deter-
mine the deontic status of a particular action from the point of view of that
domain. A moral reason to φ just is the sort of thing that counts in favor of
a moral requirement or justification to φ. This reason could be outweighed,
of course, by a stronger balance of moral reasons not to φ, and so on. With
this conceptual background in mind, the bracketing strategy is a substantive
constraint on the sort of facts that could count, as a matter of political justice
(in contrast, say, to morality), in favor of a particular action, policy, institu-
tion, and so forth. The bracketing strategy will hold that the fact that a
particular policy φ would improve the substantive welfare of the worse-off
is not, or does not provide, a political reason, or reason of justice, to φ. (I
should note here that I’m not offering a theory of what it means for some-
thing to be a reason. I am just stipulating that this is what I mean by “rea-
son.”) Instead, bracketed theories of political justice will hold that political
reasons are instead provided by, among other things, facts of what I’ll call
“non-substantive advantage”: primary goods, capabilities, or resources. In
this way, the bracketing strategy is construed as a constraint on political
theories of justice: any theory of justice that posits, as political reasons, facts
about human flourishing is ruled out.
2. POLITICAL INDEPENDENCE
Moral reasons are like political reasons. A moral reason to φ just is the sort
of thing that counts in favor of the assignment of a particular moral deontic
status to φ: justification, requirement, supererogation, etc. However, unlike
political reasons (assuming political liberalism), moral reasons are occasion-
ally provided by substantive considerations of human welfare, flourishing,
etc. The following view, for instance, seems overwhelmingly plausible:
Welfare as Reason: That φ-ing will improve the quality of life for people (or otherwise
promote their flourishing) is a moral reason to φ, which increases in strength as the
number of people whose quality of life is improved increases, and as the marginal
increase in quality of life increases.
One doesn’t have to be a consequentialist, or even a welfarist, about moral-
ity to accept Welfare as Reason. This reason could be overruled, perhaps even
trivially overruled, by a wide range of agent-centered permissions, restric-
tions, etc. But it seems entirely implausible to believe that our moral obliga-
tions should pay no attention whatsoever to the effect our actions have on
life quality. To see how plausible Welfare as Reason is, consider the following
simple example. Imagine you are in a room full of buttons. You are forced
to press some button or other (it could be any button). All other things vis-
à-vis these buttons are equal, save for the fact that one of these buttons
when pressed will substantially increase the life quality or human flourish-
ing of one individual. It would seem wildly implausible to say that one has
no moral reason to press this button. Furthermore, given that one has no
choice but to press some button or other, it would seem highly plausible to
say that this moral reason gives rise to a moral requirement: this reason is
certainly strong enough, and does not appear to be overruled by, any con-
trary moral considerations.
Notice, however, that if we accept the bracketing strategy, the fact that
this particular button increases the flourishing of some individual cannot be
a political reason to press the button. It cannot be, in other words, a reason
in favor of a requirement of justice to do so. This is not to say that there
could be no political reason to do so, especially if to do so would also be to
promote non-substantive advantage. But political reasons, given the brack-
eting strategy, are not constituted by facts about the promotion of welfare or
human flourishing. But, given the relationship between reasons and deontic
requirement (from the perspective of any domain), it would appear that if
moral reasons and political reasons tell in favor of different actions, we
should accept the following principle:
Political Independence: Political requirements may conflict with moral requirements.
One might have a requirement of political justice to φ, and a requirement of
morality not to φ.
Political Independence holds that, in certain cases, political requirements and
moral requirements will conflict. The argument for Political Independence is
relatively straightforward. Certain facts count as moral reasons that do not
also count as political reasons—including facts about the advancement of
substantive well-being or human flourishing. Thus, there are potential cir-
cumstances in which moral reasons will tell in favor of φ-ing, and political
reasons will tell in favor of ψ-ing. In such circumstances one would be
morally required to φ, and required as a matter of political justice to ψ.
Hence, Political Independence.
Actual cases of such divergence do not seem to require any conceptual
back-bending. Imagine, for instance, that a particular individual A stands in
a context in which political reasons operate. Perhaps he is deciding whether
to vote for a particular referendum; perhaps he is deciding on which policy
to support in a city council meeting, etc. Imagine that there are two policy
options, φ and ψ. Imagine that ψ is more strongly supported by the balance
of political reasons: imagine that, with respect to φ-ing, ψ-ing is favored by
the set of considerations that are significant for a bracketed conception of
94 Dale Dorsey
political justice. Imagine, however, that φ-ing will, in fact, produce a greater
level of individual well-being for all citizens. This fact, however, is not a
political reason in favor of φ-ing: given the bracketing strategy, political
reasons are not provided by the substantive facts concerning good lives for
citizens. If that’s the case, then it would appear that A has decisive political
reason to ψ, and could (at least in principle) have decisive moral reason to
φ. Hence A’s moral and political obligations would appear, in this case, to
conflict.4 (I consider a method by which to reject Political Independence in
§4. But given the divergent interests of morality and bracketed political
justice, Political Independence is a natural enough conclusion to investigate
further.)
I should note that Political Independence, by itself, should not trouble the
political liberal. After all, we can and do recognize non-moral requirements
that compete with moral requirements in individual cases. Consider, for
instance, prudential requirements. Prudence requires one to maximize one’s
self-interest. But, except on the most implausible theories of individual wel-
fare, advancing one’s own self-interest in a prudentially optimal way will
occasionally require one to perform immoral actions. Such conflict does not
entail any problems whatsoever for a first-order theory of prudence that
generates it. And so the mere fact that political and moral requirements
come apart should not by itself be regarded as a problematic result of the
bracketing strategy. Yet.
3. ALL THINGS CONSIDERED
The analogy with prudential reasons is helpful and worth spending another
paragraph on. In cases in which prudence commands us to perform some
action and morality commands us to perform some alternative action, it is
natural to ask: “What ought I to do really?” “Should I conform to moral or
prudential obligations?” This is surely an everyday feature of normative
4
One might think that I’m begging the question against political liberalism by insist-
ing that there are distinct facts about the moral point of view that might compete with
political justice. But notice that political liberals are not moral skeptics. For instance, in
a manner to be explored, Barry believes that there is such a thing as moral obligation; it
simply happens that political justice is a part of moral obligation. Furthermore, though
Rawls himself accepts a form of “Kantian Constructivism” about the moral point of
view, he does not deny that there are moral facts. He simply insists that moral facts are
suitably constructed. (See Rawls, “Kantian Constructivism in Moral Theory” in Rawls
(1999).) Hence, insofar as Welfare as Reason is decidedly plausible, any second-order
approach to moral theory should be compatible with it. Furthermore, and perhaps more
importantly, if political liberalism is committed to the denial of Welfare as Reason, this
view is a non-starter.
inquiry. This question, or set of questions, does not seek to determine either
the moral or prudential importance of performing either action. We already
know, ex hypothesi, their relative moral and prudential importance. Rather,
this question asks about the comparative practical or all-things-considered
authority of prudence and morality: in a case of conflict between prudential
and moral obligations, which has priority?
If Political Independence holds, the same question arises in the case of
political obligations. There are at least some cases in which moral and polit-
ical obligations will conflict. And, hence, it seems perfectly sensible to ask,
in any case of such divergence, whether one ought to conform to moral or
political obligations. But, or so I claim, this question causes very serious
problems for political theories that adopt the bracketing strategy.
Indeed, there are constraints on what might count as an acceptable
answer to this question, especially for the political liberal. First, any such
answer must allow, at the very least, that conforming to moral requirements
is at least all-things-considered justified. Sarah Stroud, for instance, writes:
We are inclined to take the fact that S was morally required to φ to be sufficient all-
things-considered justification for her φing, even if φing was prudentially wrong,
aesthetically unattractive, and so on. We generally accept moral necessity as sufficient
reason for φing in such cases, as we don’t for the deliverances of other evaluative
perspectives. It is as if we took the reason-giving force of moral requirements to be
strong enough to outweigh even the combined force of other evaluative perspectives
on S’s action. (Stroud 1998, 177.)
Stroud here is plumping for a principle one might identify as follows:
Permission: if x is morally required to φ at t, x has sufficient all-things-considered
reason at t to φ.
Imagine how strange it would be to deny Permission.5 Imagine that I am
morally required to perform some action, say, going out of my way to buy
a homeless man a meal. Imagine, however, that there are other countervail-
ing considerations: doing so would render me late to my son’s soccer game.
However, if we assume for the purposes of argument that one is genuinely
morally required to care for this person, it would seem terribly odd to say
that doing so is not at least justified overall. We might say, alternatively, that
other considerations might render other actions, say, showing up to the
soccer game on time, all-things-considered justified also. But it would be
strange to hold that I behaved wrongly in caring for this person given that
5
Full disclosure: some deny Permission. See, for instance, Wolf (1981); Crisp (1996).
However, the plausibility of denying Permission for Wolf and Crisp seems to stem from a
particular vision of the moral point of view, viz., an impartial vision. But, as I try to show
below, political liberalism may be stuck denying Permission for any moral view, so long as
it takes seriously Welfare as Reason—which is non-negotiable.
96 Dale Dorsey
to do so was morally required. Thus any acceptable answer to the question

of what one all-things-considered ought to do when confronted with con-
flicting moral and political obligations must at least accommodate
Permission; it must at least be permissible to conform to moral requirement.
Call this “the rationalist constraint.”
Second, any answer to the question of the rational authority of moral
and political requirements must be compatible with the most important
motivation for the bracketing strategy. In particular, it must be compati-
ble with the capacity of bracketed theories of political justice (or, indeed,
any theory of political justice) to guarantee stability for the right reasons.
Call this the “stability constraint.” To call this a “constraint” is not to say
that there couldn’t be additional motivations for a bracketed theory of
political justice going beyond stability (see, for instance, Sher (1997), chs
3–6), nor is it to say that (as discussed already) it is a decisive considera-
tion in choosing among first-order theories of political justice. But to deny
that a bracketed theory of political justice could guarantee stability is to
generate a very serious chink in the armor of bracketed theories of justice,
like political liberalism.
One might be tempted to claim that the question of stability is irrelevant
here: after all, this is a requirement to be met by theories of justice, not by
theories of all-things-considered requirements. But this reasoning is short-
sighted. The stability constraint holds that principles of justice ought to
guarantee that reasonable people will have a normally effective desire to
conform to principles of justice for the right reasons; i.e., that individuals
with fully developed moral powers, who are committed to cooperating with
others, etc., will act as the principles of justice require (on the basis of those
very principles). But whether the stability constraint holds is not simply a
matter of the first-order content of principles of political justice. It is also
a matter of the normative authority of these principles. It is a matter, in other
words, of the extent to which political principles ought to play a role in our
deliberation concerning how to act. Thus, any conception of the relative
importance of political and moral reasons must, to allow any theory of jus-
tice to accommodate the stability constraint, allow that reasonable folks will
have a normally effective desire to conform to principles of justice.
I now return to the question that introduced this section. In cases of
conflicting moral and prudential obligations, how should I, all-things-con-
sidered, act?
3.1 First Answer: Act Politically

The first obvious answer to the question of whether one should conform
to political or moral obligations is to conform to the political ones. All-
things-considered, when faced with conflicting moral and political obliga-

tions, one ought to do the just thing, rather than the morally required thing.
This proposal violates the rationalist constraint. According to the current
view, in any case of conflict between political and moral demands, one fails
to have sufficient rational justification to conform to moral obligations.
One could not, in other words, use the fact that one conforms to a moral
requirement to φ as sufficient all-things-considered justification for φ-ing.
And, hence, this view is in direct conflict with Permission.
3.2 Second Answer: Act Morally

The second option is to suggest that in a conflict between political obligations
on the one hand and moral obligations on the other, one all-things-considered
ought to conform to one’s moral requirements.
This is also a perfectly coherent option, but it violates the stability con-
straint. To see this, let me introduce a bit of terminology. Call individuals
who conform to their all-things-considered obligations “upright.” Upright
people will do what they, on balance, or as a matter of practical reasons,
ought to do. But if the answer to the question asked in this section is that
upright individuals will conform to their moral obligations in cases of con-
flict, it would appear that it is not the case that upright individuals will
“have a strong and normally effective desire to act as the principles of justice
require.” Though upright individuals may have at least some desire to main-
tain the principles of justice, this desire will be trumped by a desire to con-
form to moral obligations, a desire that will be effective when faced with
conflict cases. Furthermore, even if cases of conflict are rare, the upright
individual’s desire to conform to political requirements will remain tem-
pered: upright individuals will treat their desire to conform to political obli-
gations as secondary to the desire to conform to moral obligations, and will
hence treat the fulfillment of moral requirements as a crucial desideratum in
acting in conformity to political obligations. Insofar as these moral require-
ments will include requirements to promote substantive visions of human
flourishing, such a desideratum hardly seems compatible with the stability of
a theory of political justice that pays such visions no heed.
Of course, political liberals insist that the requirement of stability holds
of the reasonable: theories of justice should be designed to guarantee stabil-
ity among reasonable people rather than, e.g., rational or, perhaps, upright
people. (See Rawls 1995, 38–9, and elsewhere.) But the problem here seems
to me no less decisive. Let’s just assume that reasonable people will always
conform to the demands of a bracketed theory of political justice. If one
all-things-considered ought to conform to moral obligations, then it would
appear that “reasonable” individuals—those who develop an effective desire
98 Dale Dorsey
to act on the dictates of bracketed theories of justice—will not be “upright,”

as I’ve defined them. Reasonable people will not conform to the balance
of practical reasons that apply to them. But if this is right, we should ask
ourselves very serious questions about why we should believe that stability
is an attractive feature of political principles. Why should we care about
a set of political principles that will be endorsed by only those people
who don’t act as they—all things considered—ought to act? Indeed, Rawls
himself recognizes the need for political values to override others. In avoid-
ing a mere modus vivendi in a particular society, “its political values normally
outweigh whatever other values oppose them, at lest under reasonably
favorable conditions that make a constitutional democracy possible,”
(Rawls 1995, 155).
A reader suggests that the political liberal might retreat to the suggestion
that there is only pro tanto reason to bracket substantive considerations of
human flourishing. Couldn’t this proposal accommodate the proposal that
one all-things-considered ought to conform to moral demands? Perhaps, but
this proposal is importantly ambiguous. The first reading might be the sugges-
tion that the bracketing strategy provides only pro tanto political reasons—
political reasons that could be overridden by other, non-bracketed political
reasons. However, this suggestion faces obvious problems. First, it would be
a denial of political liberalism as I understand it. This view could not accept
the claim that “the state is not to do anything intended to favor or promote
any particular comprehensive doctrine.” Second, this proposal would not
guarantee stability for the right reasons given the fact of reasonable pluralism:
if substantive considerations of human flourishing can generate requirements
of justice, overriding reasons to bracket, this proposal would surely not com-
mand allegiance for the right reasons of those whose reasonable comprehen-
sive doctrines reject such an account of human flourishing. Alternatively, one
might read the proposal differently. Perhaps, instead, political reasons can be
overridden by moral reasons when it comes to all-things-considered normative
obligations. But, once again, this fails to guarantee stability, which is a central
motivation for the bracketing strategy however weak reasons to bracket are.
The problems on display here are particular to political liberalism.
Though it may be that many potential political theories imply Political
Independence, including some that will accept that facts about substantive
human flourishing can count as political reasons, it is always open to the
partisans of such theories to hold that one can permissibly ignore political
obligations when such obligations conflict with moral obligations. But this
option is not open for the political liberal: because the political liberal relies
on stability to motivate her view, she cannot then insist that it is open for
upright agents not to conform to political obligations when they conflict
with moral ones.
3.3 Third Answer: Balancing

A further possibility is to say that there is no blanket answer to the question
of what one all-things-considered ought to do in a case of conflict between
political and moral obligations. Rather, we simply weigh up the various
reasons. Sometimes, the answer may be to conform to moral requirements
(if, for instance, the political considerations are comparatively weak) or the
answer may be to conform to political requirements (if the moral consider-
ations are comparatively weak).
But this seems to me to salvage neither the rationalist nor stability con-
straints. It certainly does not salvage the rationalist constraint: in any case in
which one is rationally required to conform to political reasons, one will lack
permission to conform to moral requirements. But the rationalist constraint
holds, plausibly, that one always has permission to do one’s moral duty.
However, in any case in which the permission to conform to moral demands
is upheld, the stability of political justice is threatened. And though it is
perhaps threatened to a less substantial degree than under the suggestion that
one should always conform to moral demands, it remains the case that
upright individuals will act in an unjust way at least some of the time.
Bottom line: this proposal does not—as per the stability constraint—allow
that a theory of political justice could guarantee a normally effective desire to
act in accordance with its demands. Whether individuals will have a “nor-
mally effective desire” to act as political justice prescribes will fully depend on
the “facts on the ground,” as it were—the strength of contrary moral reasons.
This is true no matter how bracketed one’s theory of justice.
3.4 Fourth Answer: Act Either Way

If we accept Permission, and wish to hold that political obligations have at
least some practical force, we might decide to adopt a dualism of practical
reason.6 We might say, simply, that in any conflict between moral and polit-
ical reasons, one has sufficient all-things-considered reason to both conform
to one’s political obligations, and conform to one’s moral obligations. Conflict
between moral and political obligations cannot be rationally resolved one
way or the other.
Rawls tantalizingly suggests something like this picture.7 As Rawls notes
in the “Reply to Habermas,” “[i]n public reason the justification of the
6
The dualist idea is obviously inspired by Sidgwick, who could find no justification
for denying that we always have sufficient reason to conform to moral demands and the
demands of prudence. See Sidgwick (1981).
7
Thanks to an anonymous reviewer.
100 Dale Dorsey
political conception takes into account only political values . . . But since

political justification is pro tanto, it may be overridden by citizens compre-
hensive doctrines when all values are tallied up,” (Rawls 1995, 386). Further,
in discussing what Rawls calls “full justification of [the political] conception
by an individual person in society,” he states that such justification requires
only that the political conception be embedded “in some way into the citi-
zen’s comprehensive doctrine,” but that “it is left to each citizen, individu-
ally or in association with others, to say how the claims of political justice
are to be ordered, or weighed, against nonpolitical values. The political con-
ception gives no guidance in such questions, since it does not say how non-
political values are to be counted. This guidance belongs to citizens’
comprehensive doctrines,” (Rawls 1995, 386–7). Here, Rawls does not
commit to whether political or moral requirements should override, but
leaves it open for individuals to permissibly select between political and
nonpolitical, e.g., moral (given Political Independence) values.
Whether or not this is the right reading of Rawls,8 to refuse to grant
political obligations overriding normative status once again violates the sta-
bility constraint. As already explored, if we say that citizens are all-things-
considered justified in refusing to conform to obligations of justice in favor
of conforming to moral requirements it would appear that there is no guar-
antee that upright individuals will come to develop an interest in maintain-
ing and acting on bracketed principles of justice over time. Upright
individuals may, of course, choose to act on bracketed principles of justice.
But they also needn’t. And if this is correct, it is difficult to see how the cur-
rent proposal, i.e., that conforming to either moral or political obligations
could be all-things-considered justified, could allow political principles to
guarantee anything like stability for the right reasons.
The foregoing has shown, I think, a very deep problem at the heart of
political liberalism. The bracketing strategy is motivated by a commitment
to stability: reasonable (and, plausibly, “upright”) people will have a nor-
mally effective desire, for the right reasons, to accord with political justice.
Political principles should be sufficient to guarantee this form of stability.
But recall that whether or not upright individuals will endorse and desire to
8
Reason for doubt: Rawls allows that there are three “levels” of justification of a polit-
ical conception: pro-tanto justification, full justification, and “public justification by
political society.” The first and second levels of justification allow that a political concep-
tion may be justified by a given citizen, but ignored in favor of other nonpolitical values.
However, when conceptions of justice maintain only “full” justification, at best we have
a modus vivendi, not stability for the right reasons. Stability requires “public” justification,
which in turn—as Rawls explicitly notes—requires that political values override nonpo-
litical values. (Rawls 1995, 392.)
conform to political justice is a matter not simply of political principles, but

also of their all-things-considered normative authority. And hence any
account of practical rationality that allows that conforming to requirements
of justice is all-things-considered optional or, worse yet, all-things-consid-
ered overruled, will not allow any theory of political justice, no matter how
bracketed, to guarantee that justice is stable so long as that theory implies
Political Independence. But if political liberalism cannot appeal to the
requirement of stability, it has lost an essential motivation for bracketing
substantive facts of the human good.
4. POLITICAL INDEPENDENCE RECONSIDERED
If we accept the rationalist and stability constraints, political liberalism can-

not survive Political Independence. Once we accept this thesis, we must
either reject the rationalist constraint or the stability constraint. But the
political liberal could avoid this choice altogether if she could reject Political
Independence.
Recall the argument from the bracketing strategy to Political Independence:
1. That φ-ing will improve x’s welfare, well-being, or flourishing is a moral
reason to φ.
2. That φ-ing will improve x’s welfare, well-being, or flourishing is not a
political reason to φ.
3. Hence, by (1) and (2), one might have a political reason to φ and lack a
moral reason to φ, and vice versa.
4. Moral reasons determine the moral deontic status of individual actions;
political reasons determine the political deontic status of particular
actions.
5. Hence, by (3) and (4), moral and political requirements will on occasion
diverge. (And this just is Political Independence.)
The political liberal must deny a premise or reject the reasoning. But
what premises? The substantive premises are (1), (2), and (4). But (2) just is
the bracketing strategy. The denial of (4), which is a simple conceptual
truth, is also off limits. What about (1)? To deny (1), one would have to deny
Welfare as Reason. However, this strategy seems to me a clear loser. Virtually
every plausible moral theory will take some conception of welfare or human
flourishing seriously, and will adopt a positive moral stance toward the
whatever it is that constitutes a flourishing or good life.
However, even if the premises seem ironclad, the reasoning in the above
argument is weak in at least one way. The fact that moral reasons will on
102 Dale Dorsey
occasion diverge from political reasons, given that some moral reasons are
culled from an interest in the promotion of human flourishing, is not
enough to guarantee that these moral reasons won’t be overridden by other
morally important concerns. And one such concern might very well be polit-
ical justice itself. For instance, Barry writes: “Justice is, of course, a moral
concept: it is wrong to behave unjustly because that is to breach the terms
of a fair agreement for mutual constraint,” (Barry 1992, 77). Following
Barry, Rawls writes:
While a political conception of justice is, of course, a moral conception, it is a moral
conception worked out for a specific kind of subject, namely, for political, social,
and economic institutions. In particular, justice as fairness is framed to apply to
what I have called the “basic structure” of a modern constitutional democracy.
(Rawls, “Justice as Fairness: Political not Metaphysical” in Rawls (1999), 389.)
Here’s how the rejection of Political Independence, on this view, would work.
Though there are moral reasons to promote human flourishing, in the face
of a contrary requirement of political justice, these moral reasons are over-
ridden by stronger moral reasons, viz., reasons to conform to the require-
ments of political justice. This seems to follow from the suggestion that a
conception of justice is a “moral conception”: if theories of justice are, at
least in part, theories of morality, the requirements of justice, or so it would
appear, just are moral requirements.
If this proposal succeeds, Political Independence fails. But there are impor-
tant questions that must first be answered. How might we conceive of moral
reasons to conform to political justice? First, one might say that the consid-
erations that tell in favor of a political requirement to φ (assuming a brack-
eted theory of political justice) are also considerations that tell in favor of a
moral requirement to φ. This strategy simply adopts a particular account of
the moral point of view: morality cares (in a decisive way) about exactly
those considerations that (bracketed) political justice cares about. Call this
the “non-architectonic” strategy. However, one could accept the “architec-
tonic” strategy. This proposal says that, leaving aside the precise content of
political justice, there is a moral reason—indeed, a decisive moral reason—
to conform to political justice. Morality commands conformity to political
justice simply because that’s what justice requires. If either proposal is correct,
the argument for Political Independence fails, despite the truth of its prem-
ises. I take these proposals in reverse order.
4.1 The Architectonic Strategy

First, take the architectonic strategy. Is it plausible to say that one is morally
required to conform to one’s political obligations no matter what those political
obligations are, or how they are determined or formulated? Certainly not!

Whether one is morally required to conform to the demands of any particu-
lar normative system S surely waits upon whether there is a moral reason to
do that which S requires one to do. If, for instance, it is one’s political duty
to, say, bow to the nobility or to conform to a range of laws that treat one
as a second-class citizen, it is certainly not the case that one is morally
required to conform to whatever it is political reasons require one to do.
This becomes even more clear if we examine how strong this reason must
be to defend the rejection of Political Independence. To deny Political
Independence architectonically, one must hold that not only do we have a
moral reason to conform to political obligations whatever they happen to
be, but this moral reason overrides all substantive moral considerations of
human flourishing, welfare, and the like. But this is terribly implausible. We
certainly wouldn’t say that there is a moral reason with this kind of power to
conform to political justice whatever political justice happens to demand.
Of course, the natural response to this suggestion is that it is most cer-
tainly not one’s political duty to bow to the nobility or to conform to a range
of laws that treat one as a second-class citizen. All perfectly true, no doubt.
But this does nothing to salvage the architectonic strategy. The architectonic
strategy only succeeds if there is a general moral reason to conform to one’s
political obligations whatever they are. Once we hold that moral require-
ments to conform to political obligations wait upon just what those political
obligations are and whether there is a genuine moral reason to conform to those
political obligations, we reject the architectonic strategy.
4.2 The Non-Architectonic Strategy

Recall that the non-architectonic strategy holds that the facts that give rise
to political reasons (such as the fact that a particular action φ will improve
access to primary goods for the least well-off) also give rise to moral reasons:
morality takes a fundamental interest in the same sorts of facts in which a
suitably bracketed first-order theory of political justice takes an interest.9
Notice, however, to show that the non-architectonic strategy holds, one
must do so under a methodological constraint. For the non-architectonic
strategy to work, it must be that the denial of Political Independence is
implied by an independently plausible account of the content of moral
demands—independent, that is, of a question-begging precommitment
to the denial of Political Independence. Hence, the methodological constraint:
the denial of Political Independence—coincidence with a bracketed theory
9
Thanks to Steven Wall for a helpful discussion.
104 Dale Dorsey
of political justice—cannot itself be a reason to favor a particular moral

theory rather than another; this would be to abandon the non-architectonic
strategy. Let me put this another way. If to deny Political Independence one
must adopt a less plausible moral theory than one that implies Political
Independence, this just shows that the non-architectonic strategy cannot
succeed. According to the most plausible moral theory, Political Independence
fails; hence, the non-architectonic strategy fails. When it comes to the suc-
cess of the non-architectonic strategy, everything rides on the moral plausi-
bility of the theory that allows it.
But if this is correct, whether or not the non-architectonic strategy
succeeds simply comes down to a first-order inquiry into morality. Let’s
rehearse, then, just what must be the case for the non-architectonic strategy
to succeed. First, it must be the case that there are independently plausible
moral reasons that direct an individual (or an institution) to perform not
just some actions, but all actions (or policies) that are supported by brack-
eted political justice.
Just for the sake of a terminological placeholder, call these “reasons of
respect” (whatever they are or might be). But even if such reasons exist, it is
not enough to reject Political Independence to show that reasons of respect
exist and tell against actions that violate requirements of justice. It must also
be that reasons of respect bear a certain normative relationship to reasons of
welfare (or reasons derived from other substantive conceptions of human
flourishing—I’ll continue to refer to reasons of welfare here, just as, again,
a placeholder). In particular, it must be the case that reasons of respect
trump reasons of welfare. I use the term “trump” here in a technical sense.
Whenever reasons of respect tell against a certain action φ, in favor of which
reasons of welfare tell, it must be the case that in every circumstance like this,
reasons of respect are sufficient to prevent reasons of welfare from generating
a moral requirement to φ. Without this strong dominance relation, it
remains the case that reasons of welfare will tell in favor of actions that
conflict with those commanded by bracketed political justice in a way that
will sometimes generate moral requirements to violate bracketed political
principles.
So, to recap, the non-architectonic strategy runs as follows. First, per se
moral reasons of respect exist, and coincide perfectly with the obligations of
bracketed political justice. Second, moral reasons of respect always morally
trump reasons of welfare. Let’s test this view. Say that you and I have very
different conceptions of welfare. And even though my conception might be
false, it may be that morality commands you not to act in a way that pro-
motes a conception of my welfare—which may nevertheless be supported by
welfarist moral reasons—which is inconsistent with my own—false—view.
If I believe that counting blades of grass is what makes my life best, and you
act in such a way as to cause me to develop my rational capacities in a way
that does not conform to my conception of the good, though this may
actually increase my welfare, it may also violate a moral reason of respect,
given my own attitudes toward the way in which my life goes best. If the
view on the table is correct, moral obligation is itself bracketed: if a particu-
lar policy affects a set of individuals, this policy cannot be morally required
if it promotes the substantive flourishing of those individuals in ways with
which they disagree. But given that, in a political context, virtually all citi-
zens will be affected, this seems to entail that morality will not require indi-
viduals acting in a political context to promote substantive flourishing given
the fact of reasonable pluralism.
This idea has a sheen of plausibility, but I’m skeptical. Most importantly,
it seems to me that the best explanation for a moral prohibition on impos-
ing your view of welfare on my life is that my own account of what is best
for me is the best sort of evidence of what is actually good for me. We have
a tendency to believe that a person’s welfare cannot be advanced by impos-
ing a state of affairs against which this person is alienated, or which this
person does not endorse. (See, for instance, Railton, “Facts and Values” in
Railton (2003), 47.) But, just for my money, it seems implausible to believe
that morality would entail that we cannot be required to promote someone’s
welfare—no matter how much better that person’s life would go—if that
person simply has a mistaken view about the quality of their lives. Morally
speaking, such a principle sounds repugnant.
Indeed, the failure of this proposal is even more clear when we investigate
the strength reasons of respect must have. To see this in more detail, con-
sider the following case:
Troy: Ten people live in the midst of severe suffering and deprivation. To relieve this
condition, two policies are open. Policy A would alleviate the material deprivation and
suffering for all, but would do nothing more to enhance the quality of their lives. Policy
B, on the other hand, would not simply relieve their material deprivation and suffering,
but would also be to adopt a governmental policy of promoting human flourishing,
viz., the exercise and development of the capacity for human perfection, including
(most importantly) rational capacities (assume this is the right view of human
flourishing, just for the sake of argument). Nine of these people accept and value the
exercise of their rational capacities. A tenth, Troy, rejects the value of exercising his
capacity for perfection, instead preferring to maximize his own experience of pleasure.
Imagine that to adopt Policy B would be to increase the substantive flourishing of all
ten, but would very slightly decrease everyone’s achievement of pleasure and non-
substantive advantage (which is more than made up for, in terms of human flourishing,
by their substantially enhanced rational capacities). Take it as given that only Policy B
would do anything more to improve anyone’s life than Policy A.
106 Dale Dorsey
Here, reasons of respect tell very clearly against Policy B. The only reasons
that would motivate Policy B rather than Policy A are those that refer explic-
itly to a particular theory of human flourishing, viz., that to flourish one
should exercise one’s rational capacities. To adopt Policy B violates a reason
of respect: it is to act in such a way as to promote a substantive vision of
Troy’s flourishing or welfare that he rejects. And, hence, if reasons of respect
trump reasons of welfare, Policy A is commanded.
Strictly as a claim about morality, this is simply implausible. To adopt
Policy B would be to improve the lives not just of Troy (despite his own
judgment) but also of nine other people. Given this, however, it seems absurd
to say that Troy’s own false belief about his flourishing would or should
stand in the way of a moral requirement to genuinely improve the lives of all
ten. Imagine that it’s not Troy-versus-nine, but Troy-versus-ninety-nine, or
Troy-versus-999, etc. At some point it begins to look mighty implausible to
say that one is not morally required to act in accordance with reasons of
welfare rather than reasons of respect. Considered simply as a proposal
internal to the moral point of view, holding that reasons of respect trump
reasons of welfare does not pass the smell test.
Troy’s case should convince us of the following: reasons of respect, con-
strued in their most plausible formulation, do not trump reasons of welfare.
There is some point at which reasons of welfare are sufficient to override
reasons not to advance someone’s welfare (i.e., Troy’s) in a way he or she
rejects. And, hence, a crucial supposition upon which the non-architectonic
strategy rests cannot succeed.
I’ll consider three responses here. First, one might think that the sheer
weight of the good lives for the people in question might be enough for
moral reasons of respect to favor Policy B. On this view, reasons of respect
do not rule out actions that promote a particular vision of human flourish-
ing for those who reject this vision so long as sufficient welfare is generated
by the action in question. This may be plausible, but it is no savior for
political liberals who would deny Political Independence. If it’s true, then
reasons of respect cannot always coincide with the demands of bracketed
political justice. It is hard to see how the integrity of any bracketed political
theory could be maintained in light of the suggestion that the promotion of
a substantive vision human welfare is perfectly acceptable as an explicit aim
of political justice so long as enough welfare is promoted by such policies.
Here’s a twist on this suggestion. Perhaps reasons of respect do not count
against a particular policy or action that promotes a substantive vision of
human flourishing for a person that said person rejects if said person
would also assent to the policy despite his or her rejection of this vision of
human flourishing. We might assume that Troy himself is not just a rea-
sonable but a beneficent sort: he’s willing to accept a policy that would
promote a particular conception of welfare not his own if it’s the case that
in so doing many others conform to a life they believe is worthwhile. So,
in other words, though he does not endorse the value of his rational capac-
ities per se, he does endorse the plan’s proposal given the attitudes of others.
And hence, or so it may be claimed, there is no reason of respect to refrain
from benefiting Troy in this case.
I think there are two problems with this suggestion. The first is that Troy’s
assent to Policy B does not seem to pass the test of stability: given the case
as imagined Troy does not consent to Policy B for the right reasons, but
instead simply given its instrumental benefits. But leaving this aside, the
biggest problem with this suggestion is that even if in Troy’s case, given Troy’s
dispositions, there is no reason of respect not to advance Policy B, not all
people have Troy’s attitudes. Troy could reasonably take a different line. Troy
could refuse to grant Policy B his support given that he believes that it will
do nothing to advance the flourishing of anyone, would be to grant sub-
stantive governmental support to a conception of the good that is not his
own, would cost him a modicum of pleasure, which he values, and would
reduce his non-substantive advantage. And, hence, even if Troy himself
were to grant support to the policy, this would do nothing to guarantee that
reasons of respect, even if such support were relevant, would plausibly
trump moral reasons in other cases.
Third, and finally, one might hold that reasons of respect—construed as
reasons against the promotion of a substantive vision of human flourishing
for individual agents that they reject—are not the only reasons that may
work to yield a denial of Political Independence. In other words, it may be
that not just reasons of respect, but many other potential reasons as well,
have moral weight, yield moral reasons to act as bracketed political justice
requires, and (when combined) can trump reasons of welfare. But Troy’s
case sheds doubt on any such suggestion. This is because Policy A is very
clearly superior when it comes to those things that bracketed political jus-
tice cares about—including non-substantive advantage—and inferior only
when it comes to those things that bracketed political justice very clearly
doesn’t care about, viz., substantive human flourishing. So, however one
understands the moral reasons that command us to act as bracketed justice
requires, no such account of these reasons will plausibly trump reasons of
welfare, as seen in Troy’s case. Morally speaking, Policy B is required. To say
otherwise is to adopt a distorted vision of the moral point of view simply for
the sake of denying Political Independence, and as such can do nothing to
support the non-architectonic strategy.
In sum, I think Troy’s case sheds substantial doubt on the non-architec-
tonic strategy no matter how one seeks to accommodate it. That cases like
Troy’s exist should come as no surprise. To reject Political Independence, the
108 Dale Dorsey
political liberal must thread an extremely fine needle: she must show that
there are no cases in which an independently plausible moral theory would
require us to act contrary to a bracketed theory of political justice. But to do
this she must not only argue that there are independent reasons to accept
a set of moral reasons that tell exactly in favor of those actions supported
by bracketed political justice—which itself would seem cosmically coinci-
dental, but which I have granted for the sake of argument—and that these
reasons, whatever they are, have the moral power to override reasons of
welfare in every possible case. But this is an extraordinarily strong thesis
independently of a substantive consideration of its plausibility. And hence
that there are counterexamples, potentially a number of counterexamples,
should not be difficult to swallow.
Given the failure both of the architectonic and non-architectonic strate-
gies, it would appear that Political Independence (on the assumption of a
bracketed theory of political justice) is well supported. But this leaves the
political liberal in exactly the pickle I presented at the conclusion of §3:
the political liberal, because she must accept the stability constraint, and can-
not deny Political Independence plausibly, must reject the rationalist constraint.
But to reject the rationalist constraint is to deny not just a very plausible prin-
ciple of the relationship between morality and how we ought to live, it is to
deny a principle that most look upon as little more than platitudinous. I claim,
therefore, that the best option is to reject the stability constraint.
5. BEYOND STABILITY
In this chapter I’ve explored three mutually incompatible ideas, viz., Political
Independence, the stability constraint, and the rationalist constraint. A brack-
eted theory of justice is stuck with Political Independence. A denial of the
rationalist constraint is implausible. And, hence, it would seem like the best
option is to deny the stability constraint. However, one might question
whether this is really fatal to political liberalism, as I understand it here.
Couldn’t there be other motivations for a bracketed theory of political justice?
Ultimately, I do not rule out the possibility that political liberalism could
survive the critique I offer here by outlining an alternative rationale and
instead simply abandoning the importance of stability for the right reasons.
I leave this as an officially open possibility. But I think that Political
Independence, combined with the rationalist constraint, yields a very impor-
tant challenge for any bracketed theory of political justice, even if we don’t
accept stability as the primary rationale. If we accept Political Independence
and Permission, this entails that it is perfectly permissible—compatible with
all-things-considered obligations—for people, including those who set the
terms of public policy and the rules enforced by state apparatus, to design
political institutions and public policy in accordance with moral obliga-
tions, which refuse to bracket substantive considerations of well-being. But
if lawmakers and designers of the basic institutions of society act in a per-
fectly justified manner in promoting a particular vision of human flourish-
ing, a rationale for political liberalism must offer an explanation of why it
would be impermissible for political agents to do so simply on political
grounds.
But this explanatory burden, or so I claim, is heavy indeed. The weak-
ened normative authority of political justice in fact wrecks a number of
possible rationales for a bracketed theory: Political Independence and
Permission entail that even a bracketed theory of political justice will not
help to satisfy a desire “to live in a society whose members all freely accept
its rules of justice and its major institutions,” (Barry 1992, 164) nor does it
guarantee that political institutions will be compatible with the full exercise
of autonomy (Sher 1997, chs 3–4),10 nor does it guarantee that “citizens will
want to propose and abide by” the rules of justice (Quong 2011, 143), if any
of these ideas are construed in a manner that would support a bracketed
theory of political justice. (Of course, one could construe the importance of
agreement, say, as compatible with political institutions being guided by
moral requirements on occasion—but given the influence human flourish-
ing has over moral requirements, this rationale would obviously not support
political liberalism as I construe it here.)
So while I leave open the possibility that there may be some additional
rationale to support bracketing, any such rationale must be able to preserve
the importance of accepting a bracketed theory of justice in light of the fact
that political actors do not necessarily act wrongly in designing their insti-
tutions in accordance with considerations of substantive human flourish-
ing. In my view, we have reason to be skeptical of this possibility, but I’ll not
press the point further.
6. CONCLUSION
The bracketing strategy seems unable to plausibly avoid Political Independence.

And so it would appear that any partisan of the bracketing strategy, including
10
Of course, this is controversial insofar as (a) the “autonomy” rationale seems incom-
patible with the bracketing strategy as I understand it, insofar as it treats the promotion
of substantive autonomy as a political reason and (b) Sher argues, plausibly, that even an
appeal to autonomy cannot justify bracketing substantive considerations of the human
good when it comes to public policy.
110 Dale Dorsey
so-called political liberals, must either reject this strategy or reject the
rationalist constraint. There is no feasible method by which to retain the
rationalist constraint and also maintain the fundamental motivation for
the bracketing strategy, i.e., the insistence on stability of political principles
for the right reasons.
As I’ve noted throughout, there are ways that political liberals might
respond to the argument I offer here that preserve their position. First, they
might accept a stronger form of moral anti-rationalism, and deny Permission.
Second, they might take up the project of offering an independently plausi-
ble (independent, that is, of whether Political Independence is true or not)
account of the moral point of view with an eye to demonstrating that Political
Independence should be rejected despite Welfare as Reason. Alternatively, and
more radically, one could deny the stability constraint, hold that the stability
of a system of political justice need only be guaranteed perhaps in terms of a
modus vivendi, and instead offer some other motivation for a bracketed the-
ory of political justice. Each of these proposals has significant challenges (the
first a challenge of plausibility, the second a challenge of, as it were, “high
odds,” the third the challenge of outlining a rationale that is unaffected by
the fact that political institutions will be designed by upright people in
accordance with facts of human flourishing even if political justice says they
should not do so). In any event, while there are certainly options available for
the political liberal, I have tried to bring out the essential challenges that this
view faces in light of the competing authority of moral and political demands.
Bibliography
Barry, Brian (1992). Justice as Impartiality. Oxford: Oxford University Press.
Crisp, Roger (1996). “The Dualism of Practical Reason” in Proceedings of the
Aristotelian Society, 96.
Quong, Jonathan (2011). Liberalism without Perfection. Oxford: Oxford University
Press.
Railton, Peter (2003). Facts, Values, and Norms. Cambridge: Cambridge University
Press.
Rawls, John (1971). A Theory of Justice. Cambridge, MA: Harvard University Press.
Rawls, John (1995). Political Liberalism. New York, NY: Columbia University Press.
Rawls, John (1999). Collected Papers, ed. Freeman. Cambridge, MA: Harvard
University Press.
Sandel, Michael (1989). “Moral Argument and Liberal Toleration: Abortion and
Homosexuality” in California Law Review, 77.
Sher, George (1997). Beyond Neutrality. Cambridge: Cambridge University Press.
Sidgwick, Henry (1981) [1907]. The Methods of Ethics. Indianapolis, IN: Hackett
Publishing Company, 7th edn.
Stroud, Sarah (1998). “Moral Overridingness and Moral Theory” in Pacific

Philosophical Quarterly, 79.
Wall, Steven (1998). Liberalism, Perfectionism, and Restraint. Cambridge: Cambridge
University Press.
Wolf, Susan (1981). “Moral Saints” in Journal of Philosophy, 79.
5
Against Public Reason
David Enoch
1. BY WAY OF INTRODUCTION: FRUSTRATION
I don’t know of any other philosophical discussion that is quite like this: talk
to (broadly speaking) Rawlsians, and you are likely to get the impression
that some kind of political liberalism, or a public-reason account, is the only
game in town.* The only questions worth discussing, it seems, are within
this framework, rather than about it. Theorists who reject this framework
are often ignored, and the feeling one gets is that they just don’t get it. Talk
to many others (non-Rawlsian political philosophers, as well as philoso-
phers whose main area is not in political philosophy exactly), and you are
likely to get the impression that Rawlsian public reason has been effectively
refuted several times over and, indeed, that even this much was never nec-
essary, as the theory was a non-starter to begin with. And the feeling one
gets is that Rawlsians just don’t get it, and that their tremendous influence in
political philosophy is corrupting the field.
As my title suggests, I am much closer to the second stance than to the
first (though both have been presented here in rather extreme, somewhat
caricatured, forms). So, in this chapter I make yet another attempt to explain
why public-reason accounts are hopeless. I share the feeling that there are
already in the literature strong reasons to reject specific public-reason
accounts and perhaps that tradition more generally, but I think that more
can be done—and that given the state of the field, more should be done.
In this chapter, then, I try to develop some of the strongest, most general
* Earlier versions of this chapter were presented at the ANU, Aarhus University
Copenhagen, Leeds, Arizona, Yale, CUNY, Rutgers, Tel Aviv, and Jerusalem. I thank the
participants on those occasions for the discussion and their valuable comments. And for
extremely helpful written comments, I would like to thank Dani Attas, Dan Baras, Ilan
Ben Shalom, Steve Darwall, David Estlund, Ruth Gavison, Alon Harel, Shelly Kagan,
Barak Medina, Shmulik Nili, Shai Perry, Jon Quong, Chad van Schoelant, Steve Wall,
Alex Zakaras, and an anonymous reader for Oxford Studies in Political Philosophy.
Against Public Reason 113
objections to public-reason accounts.1 For this reason, I do not focus on the

details of specific public-reason accounts,2 not even Rawls’s. Furthermore,
in an attempt to get past the they-just-don’t-get-it predicament, I try to
broaden my vision, and to place some of the controversies here in a wider
philosophical context. Also, I try to pinpoint—from the point of view
of a public-reason-skeptic—the underlying intuition that public-reason
accounts get right, and to accommodate it without following them to their
(unacceptable, I argue) conclusions.
In my more optimistic moments, then, I hope that this chapter not
only presents sound arguments, but also that it will advance the debate. But
I am not unrealistic: I fully expect—in the descriptive, not the normative,
sense—to come up against he-just-doesn’t-get-it responses.
At the end of the next section, after presenting what I take to be the main
underlying intuitions that push people in the direction of public-reason
accounts, I present a general characterization of the public-reason accounts
that will be my target in the rest of the chapter. Public-reason accounts, as
I characterize them, are the set of views that are best seen as motivated by
these or closely related intuitions, and that as a result endorse some require-
ment to justify political principles3 to each of those subject to them as a
necessary condition for legitimacy. I then proceed, in section 3, to note that
all public-reason accounts must involve some idealization—because there is
pretty much nothing citizens in modern societies all agree on, the require-
ment to justify political action to all requires (if anarchism is to be avoided)
either restricting the scope of those to whom justification is owed, or idealiz-
ing the conditions under which such justification is owed, or both. General
thoughts about when such idealization is a legitimate theoretical move and
when it’s objectionably ad hoc—as well as reflection about the specific ideal-
ization involved in typical public-reason accounts—show that both options
are subject to devastating critiques. In section 4, I defuse another one of the
motivating thoughts of public-reason accounts by discussing what reason
we act on when we impose on others a principle we (but not they) believe in.
I insist that the relevant reason for action is the content of the principle, not
1
Two I don’t address here: self-defeat (which I address, following others to an extent,
elsewhere; see Wall (2002), and my (2013)); and a critical discussion of the epistemic
commitments of public-reason theorists, which I intend to address in future work.
2
I do some of this elsewhere, with regard to Estlund (2008; see my (2009a)), Gaus
(2011; see my (2013)), and to an extent Waldron (1999; see my (2007)). (In what follows,
from time to time I borrow a few sentences from these other texts of mine.) Almost all
other critics of this tradition target primarily Rawls, of course.
3
For my purposes here it won’t be necessary to distinguish between public-reason
theorists who (following Rawls) restrict the scope of this requirement to just, roughly
speaking, constitutional essentials, and those who (like Quong) apply the requirement
more broadly.
114 David Enoch
that we accept it. This means that the impression sometimes highlighted by
public-reason theorists that by flouting a public justifiability requirement
we unacceptably give our own views extra weight is confused. In section 5,
I take a step back from the details of some of the earlier sections, reflecting
on the role of the political philosopher according to public-reason theorists.
I highlight what I take to be a deep, dangerous mistake—among other
things, this conception of the role of political philosophers divorces them
entirely from even the just struggles of political activists. In the concluding
section I return to the starting point—to the powerful underlying intuition
that public-reason theorists, I think, get right. I show how such intuitions
can be accommodated without endorsing a public-reason account.
2. THE UNDERLYING PUBLIC REASON INTUITIONS
Suppose that Catholicism is the true religion, and that the Pope has a direct
line to God. Still, if you are a non-believer (or a believer of some other
religion), we find it deeply objectionable for the Pope’s directives to have
authority over you, simply in virtue of the truth of Catholicism and the Pope’s
excellent epistemic credentials.4 Certainly, the thought that you can be
politically subject to the Pope’s directives—backed up by the state’s coercive
power—seems unacceptable. The reason, it seems, is that while his direc-
tives are in some sense justified, they are not justified to you, non-Catholic
as you are. And in order to reconcile legitimate authority with your nature
as a free, autonomous agent, there must be something to be said for it, and
furthermore, this something must be sayable to you. Truth, it seems, is never
sufficient for legitimacy. Rather, those over whom authority is claimed are
entitled to answers, to justifications of the authority that are available to
them, pretty much as they are.
The general thought seems to be that given the liberty, autonomy, ration-
ality, or some such of those subject to the purported authority, its legitimacy
must be somehow accessible to them, it must be such that they too can
appreciate it, it must be such that they have reason to endorse it, or, as the
slogan goes, it must be justifiable to each and every one of them. It is impos-
sible, the thought seems to be, for a free person to be subject to a legitimate
authority if there’s nothing that can be said to him, in terms he can relate to,
as it were, that justifies subjecting him to that authority. Notice that no
consent is required—perhaps, for instance, it’s not about engaging your will
(as consent presumably is) but your reason (a central distinction in Waldron
4
Estlund’s (2008, 5) example.
(1987)).5 Still, the natural thought is that unless an authority can be justified
to you pretty much as you are, it does not have legitimacy over you.
For brevity, I am going to put this underlying thought as the attempt to
reconcile authority with liberty. Of course, these terms (“authority” and
“liberty”) are neither univocal nor philosophically transparent.6 But for my
purposes here the somewhat loose characterization above suffices—clearly,
something like the tension between authority and liberty does serious work
here, as otherwise it would be entirely mysterious why we should justify an
authority to those subject to it rather than to others. It is they who are owed
justification, because it is their liberty (or some such) that is at stake when
they are subjected to an authority. In my criticism of public reason below,
nowhere will I rely on a more specific, more controversial characterization
of the notion of liberty7 or of the need to reconcile liberty and authority.
Furthermore—and this brings us to the second underlying public-reason
intuition—for the Pope (or any of his followers) to apply his Catholic-based
directives to you non-believer as you are, will amount to giving more polit-
ical weight to his (or their) beliefs than to yours, and so will amount not
only to failing to treat you as free, but also to failing to treat you as an equal
citizen, one whose beliefs, principles, desires, projects are as politically sig-
nificant as anyone else’s. When we combine these two underlying intui-
tions, we get the slogan—the justifiability-to requirement follows from the
commitment to treat each other as free and equal.
This, then, is what in this chapter I take to be characteristic of public-reason
theories. They are all committed to some kind of accessibility requirement
as a necessary condition for legitimacy: the thought that for a state (for
instance) to be legitimate, its authority must be justifiable to each of those
subject to it. And—and this will be important in what follows—they all
share the underlying motivation of reconciling liberty and equality with
authority. To an extent, you should feel free to treat this characterization as
a stipulation—in what follows, I will use the words “public-reason accounts”
to designate theories that are committed to some justification-to condition,
motivated by the need to reconcile authority with liberty and equality. But
5
In fact, we can think about consent as a particular instance of this more general,
arguably more accurate, condition. For if you’ve consented to an authority, we are guar-
anteed to have something to say to you about its legitimacy: namely, that you’ve given
your consent. According to this line of thought, then, the problem with consent theories
is simply that they over-generalize.
6
For my take on authority in general, see my “Authority and Reason-Giving”
(2014).
7
In particular, I will not be assuming a general freedom-to-do-as-one-chooses con-
ception of liberty.
116 David Enoch
I don’t think that this way of using words is at all idiosyncratic, and I think
that this characterization nicely captures pretty much all and only accounts
that are usually thought of as members of this tradition. Here, for instance,
is Nagel (1991, 33–6):8
The task of discovering the conditions of legitimacy is traditionally conceived as
that of finding a way to justify a political system to everyone who is required to live
under it . . . the search for legitimacy can be thought of as an attempt to realise some
of the values of voluntary participation in a system of institutions that is unavoidably
compulsory.
And similar characterizations of the public-reason account and underlying
motivations are common.9
Now, as noted above, my hope is to criticize the entire public-reason
tradition, not any specific members thereof. So I cannot rely on more
specific details—say, the exact nature of the accessibility or justification-to
requirement.10 Though such differences may be important in other con-
texts, for my purposes here what’s important is what such different accounts
have in common. In particular, then, I can afford not to say much more
about what exactly it is that is needed for it to be the case that something is
not just justified, but is justified to the relevant constituency. Still, one thing
must be emphasized here: though I am going to remain neutral on the
precise nature of that extra thing that is needed (for justification-to), I am
going to assume that this requirement is non-vacuous—that is, that a justi-
fication that is sound in an important sense can be accessible (in the relevant
sense) to some but not to others, that the justification-to requirement takes
8
Quoted in a similar context in Rossi (2014).
9
For statements of an accessibility or justification-to requirement as the definitive
feature of this tradition, see, for instance, Waldron (1987, 128); Quong (2011, 161);
Quong (2013, the opening line); Gaus’s (2011, 263) “Basic Principle of Public
Justification;” Vallier and D’Agostino’s (2013) Public Justification Principle; and Wall
(2002, 385, 387).
For a helpful survey of possible motivations for public-reason accounts, see Quong
(2013, section 1), and the references there. There, thoughts about coercion constitute
only one section out of five. But it is, I believe, the central one, and much of what goes
on in the other sections also tacitly relies on it. For instance, another motivation Quong
brings from Gaus has to do with Gaus’s account of our reactive attitudes, and some acces-
sibility being necessary for their appropriateness. But for Gaus, the appropriateness of the
reactive attitudes is closely tied to thoughts about when the relevant part of morality is
objectionably authoritarian—that is, to an extent, inconsistent with the liberty and
autonomy of those subject to it.
10
For a helpful list of relevant distinguishing questions here (the nature of the accessibil-
ity, the constituency, the nature of the thing to be justified, etc.) and for many references,
see Vallier and D’Agostino, section 2.
as input something about the actual features of the relevant addressee.11 For
now, this will do.12
3. IDEALIZATION
I start my discussion of the role of idealization in public-reason accounts

by arguing that all (non-anarchist) public-reason accounts need to employ
some idealization (3.1). I then have an interlude on when in general ide-
alization is a legitimate philosophical move (3.2). With a general answer
to this question at hand, I proceed to reject the two idealization devices
employed by public-reason accounts—excluding the unreasonable (3.3)
and going hypothetical (3.4). Similar objections, I conjecture, will apply
to any other idealization device employed in order to save public-reason
accounts.13
3.1 Why No Public-Reason Account Can Do without It

Recall the characteristic feature of public-reason accounts—they require
that the relevant authority or principles be justified to all those subject to the
authority. But this creates a problem, at least in the context of hoping to
vindicate some contemporary states. The problem is that actual citizens of
actual large-scale contemporary states are a very varied bunch. Different
11
Raz (1998) claims that the only accessibility requirement that makes sense here is
one that is always trivially met, because all good normative arguments are always accessi-
ble to all. I don’t agree with Raz on this—I think that there is sense to be made of the
thought that the Pope’s justifications are not accessible to a non-believer, even under the
assumption that Catholicism is true. But my point in the text is just that an account that
superficially uses the justification-to lingo, but which happily takes on board this point
from Raz, is not within the target of my arguments in this chapter (nor is it naturally
classified as a public-reason account).
12
The words “public reason” are sometimes used in a different—narrower, perhaps
even more precise—sense, one having to do with the shared language of officials and the
Rawlsian duty of civility not to rely on private conceptions of justice. See Quong (2011,
41–2), and the references there. And for the distinction between the two ways of using
“public reason,” see Quong (2011, 256). I have nothing here to say about this other sense
of “public reason”—I just note it here in order to preempt confusion.
13
Indeed, arguably the nature of the idealizing device doesn’t much matter for the
points to come. In particular, there is structural similarity between excluding the unrea-
sonable and going hypothetical, perhaps to the point that a more theoretically parsimo-
nious public-reason theory will make do with just one of those (say, packing enough into
the notion of the “reasonable” so that once the constituency has been restricted to the
reasonable, no further hypotheticalization is needed). I discuss these two mechanisms
mostly because public-reason theorists employ both, usually in tandem. It should come
as no surprise that similar worries arise regarding both.
118 David Enoch
people are committed—sometimes even in the deepest ways—to all sorts of

views and doctrines, they value—even intrinsically—all sorts of different
things. If the justifications offered to them are to engage them as they actu-
ally are—perhaps based on principles they accept, or on the values they
hold dear, or on what is already there in their motivational set—then it’s
hard to believe that there is anything at all that can be justified to all. This is
perhaps clearest on consensus-versions of public-reason accounts, according
to which for a political principle (e.g.) to be legitimate there must be a jus-
tification for it that is available (in the relevant way) to all.14 But it remains
true even on convergence views, according to which the condition necessary
for legitimacy is just that for any citizen, there’s a justification available to
her (without the further requirement that it must be the very same justifica-
tion that’s available to all).15 So long as the justification-to requirement is
non-vacuous, and so long as the relevant constituency consists of all the
citizens of a contemporary state as we actually find them, it’s hard to imagine
anything at all passing the bar. Anarchism follows.
Since public-reason theorists are not anarchists, they idealize, thereby
making the justification-to requirement easier to meet. The two ways of
doing so that are used by public-reason theorists are, first, restricting the
scope of the relevant constituency (those to whom public justification is
owed as a necessary condition for legitimacy)—so that what’s needed for
legitimacy is not justifiability to all, but, say, to all the reasonable, or all the
qualified;16 and second, going hypothetical, so that those engaged by the
justification-to requirement are not people as they actually are, but some
hypothetical, idealized version thereof—perhaps, for instance, what’s
important is not that people see the justification of the relevant principle,
but that they would see or endorse it, if they spent a respectable amount of
good reasoning on it.17
3.2 Interlude: When is Idealization an Acceptable

Philosophical Move?
A theory is offered, one that ties some phenomenon to our relevant responses.
Perhaps, for instance, a theory is offered about the relations between (phe-
nomenal) color and our color appearances, so that to be red is just to appear
red. Or perhaps a theory of values is offered that ties them very closely to
14
See, for instance, Larmore (2003); Quong (2011).
15
See, for instance, Gaus (2011).
16
Rawls, and following him pretty much everyone else, talks of reasonableness. Talk
of qualified acceptance comes from Estlund (2008).
17
See Gaus (2011, 250).
what we value, so that to be of value just is to be valued. But counterexam-

ples immediately come up: sometimes, something appears red to me even
though it isn’t, and sometimes it doesn’t even though it is. Sometimes, peo-
ple value things that aren’t of value, and fail to value things that are. A nat-
ural move then is to idealize: perhaps to be red is not to appear red, but
rather to appear red to normal observers, in good lighting conditions. Or per-
haps to be of value isn’t to be valued, but to be valued by the right people, in
the right conditions.
There is something perplexing about such a move. If you are really thirsty,
but can’t find any water, offering you some ideal or hypothetical water
(water that is not available, but would have been available, if . . . ) will not be
a good substitute.18 Why are hypothetical or idealized responses any better?
The answer is that whether they are depends on what the underlying
motivation was for going for the relevant view in the first place. Think again
about the water example. Presumably, your reason for looking for water is
that it would quench your thirst. Hypothetical water won’t do that—it
won’t even go a part of the way towards doing that. So in the absence of
water, going for hypothetical water amounts to cheating—it is disconnected
from the underlying motivations of looking for water, and therefore offers
no satisfaction, not even partially.
What about colors? Is idealizing here a way of cheating, avoiding coun-
terexamples in an ad hoc way? Or are ideal, hypothetical responses enough
here? The answer depends on the philosophical motivations underlying the
relevant account of colors. If they are all about actual observers and their
actual responses, then going hypothetical (because of the pressure from
counterexamples) is cheating. What counterexamples show us, in such a
case, is not that we should settle for observations in hypothetical conditions,
but that we should resist the attempt to tie colors and our appearances of
colors as closely together as all that. But if the underlying motivations of
such an account are consistent with settling for hypothetical conditions,
then all is well, and the idealized response-dependence account may still be
a good idea.
Similarly for values and what we value. If the underlying motivations
for offering a response-dependence view of values is tied to actual people
and their responses, then what we should do in the face of the obvious coun-
terexamples is not idealize (in an ad hoc way), but rather reject response-
dependence altogether. But if the idealization can be motivated in a way
that’s consistent with going response-dependence and with the philosophi-
cal motivations for so going, then all may still be well.
18
The line here echoes a well-known criticism of hypothetical consent theory, for
instance, in our context, Dworkin (1973).
120 David Enoch
And so, we have a test for when idealization is a legitimate philosophical

move. It is, when it is motivated, and furthermore, when the offered moti-
vation is consistent with the motivations for going for the initial, non-
idealized view (the one that was devastated by obvious counterexamples). This
is why hypothetical water is out, why (perhaps) some idealized response-
dependence views of colors may be in, and why (as I argue elsewhere) ide-
alized response-dependence views of normative concepts are out.19
Now we can put the test to use with regard to the idealization that is part
and parcel of all public-reason accounts. Is it legitimate, motivated, ideali-
zation? Is it consistent with the underlying public-reason motivations? Or is
it ad hoc, the kind of idealization whose sole purpose is to artificially save a
theory from obvious counterexamples?
3.3 Excluding the Unreasonable20

Think, then, about the restriction of the justifiability-to requirement to just
the reasonable (or some such), so that the fact that a principle is not justifi-
able to the unreasonable does not undermine its legitimacy. What is the
motivation for this restriction? Can it be reconciled with the underlying
motivation for public-reason accounts?
Though what is meant by “reasonable” (even just by Rawls) is a matter of
controversy I’d rather not enter, still I think we can say, for our purposes
here, that public-reason theorists typically think of reasonableness as con-
sisting of a motivational and an epistemic element. The epistemic element
is acknowledging things like “the circumstances of justice,” reasonable plu-
ralism, or the burdens of judgment and their effects—according to such
conditions, if you just expect (in the descriptive sense) everyone to converge
on your religion, say, you’re not being reasonable. The motivational element
is roughly about wanting to interact with others as free and equal, perhaps
even being more strongly motivated by such a desire than by many others.21
If you don’t think of others as your equals for political purposes, if you’re
entirely happy interacting with them by manipulating them rather than by
reaching agreements, then you’re not reasonable.
19
See my “Why Idealize” (2005). But see also Sobel (2009) and my (manuscript).
20
Worries in this vicinity have been around since very early on. See, for instance,
Hampton (1993, 299), and Raz (1998, for instance 37). In what follows, I try to address
them both more generally (not just vis-à-vis Rawls), and in a more principled way
(based on the general thoughts in the text about when an idealization is philosophically
acceptable).
21
See, for instance, Quong’s (2011, 291) understanding of (Rawls’s understanding of )
the reasonable.
Notice that “reasonableness” here is a technical term, introduced to serve

a certain role in the theory by stipulation. Sure, it may bear some resem-
blance to how we use the term in natural language, but the characteriza-
tions of reasonableness (for instance, in the previous paragraph) are a
matter of stipulation, not of conceptual analysis.22 So—their many
attempts to the contrary notwithstanding23—those invoking the concept
cannot rely on intuitions we supposedly have about who is and who isn’t
reasonable. We don’t have such intuitions about technical terms. This is
not just a methodological glitch. For often, perhaps by relying on some
pre-theoretical intuitions about a non-technical concept of reasonableness,
public-reason theorists write as if the unreasonable are pretty much the
Nazis and the murderous psychopaths.24 Those, of course, are unreasonable
also in the natural-language meaning of the term; and frankly, excluding
them doesn’t hurt that much, and perhaps—though I’m not as sure—is not
theoretically problematic either. But this is cheating. While the Nazis and
the murderous psychopaths are obviously unreasonable, so are—on public-
reason accounts—many, many others. For instance, at least some of the
Rawlsian thoughts about the burdens of judgment and their epistemic sig-
nificance would be rejected by pretty much any epistemologist working in
the field today.25 Perhaps more importantly, if you are not a Rawlsian—if
you don’t accept something like a public-reason justification-to requirement
as a necessary condition for legitimacy—then this too suffices, according
to public-reason accounts—to make you unreasonable.26 And—perhaps
depending on the details—this may be so even if you do go for a pub-
lic-reason account, just not the right one.27 So, it’s really important to keep
in mind—when discussing the status of the unreasonable—that we’re not
talking about just the Nazis and the murderous psychopaths. For any public-
reason theorist, the ones excluded as unreasonable—that is, the ones justi-
fication to whom is not necessary for legitimacy—arguably include also
22
In this respect, Estlund’s insistence on going for a more technical term (“qualified
acceptability”)—motivated precisely by the kind of considerations in the text here—is
more helpful.
23
For an especially clear example, see Quong (2011, 156).
24
Quong’s examples, for instance, are those making “claims to racial, gender, or ethnic
superiority” (292), the Ku Klux Klan (299), white supremacists (309), Nazis (309), and
psychopaths (314).
25
I address the shaky epistemological-sounding claims made by Rawlsians in work in
progress, provisionally entitled “Political Philosophy And Epistemology: The Case of
Public Reason.”
26
For explicit claims to this effect, see Dreben (2003, 326); Estlund (2008, 61);
Quong (235, fn 34; 240). As far as I know, Rawls nowhere says this explicitly. Nor do
I know of a place where he rejects this claim.
27
For instance, I think—but I’m not entirely sure—that according to Quong’s con-
sensus view, a Gaus-like convergence view is unreasonable.
122 David Enoch
such people as John Stuart Mill, Karl Marx, Joseph Raz, Jean Hampton,
pretty much all contemporary epistemologists, probably most of those
offering a rival public-reason account, presumably the early Rawls—oh,
and me.28
Now, a case can be made that these aspects of the delineation of the
unreasonable are a necessary feature of public-reason accounts.29 But, loyal
to my attempt to target the tradition in general rather than some (even all
developed) specific manifestations thereof, let me not commit myself to this
strong claim. So, the thing to keep in mind is not necessarily that the rea-
sonable include all non-Rawlsians, but that they include many more than
just the Nazis and the murderous psychopaths.30
Can public-reason theorists retreat to a pre-theoretical understanding of
the reasonable, perhaps so that you qualify as reasonable if your reasoning
mechanisms are functioning well, or if you’re willing to listen to other views,
and so on?31 They can, of course, but the price would be anarchism again—
for among those who are reasonable in just some very thin sense of this
kind, everything is controversial. Under any plausible hypotheticalization,
and any non-trivial understanding of accessibility, nothing is justifiable to all
the reasonable in this sense.
Keeping in mind, then, how easy it is to be unreasonable, let’s return to
the philosophical motivation underlying public-reason accounts—the need
28
Public reason theorists can insist (as Quong (2011, Chapter 10) explicitly does) that
the unreasonable should enjoy the same rights that the reasonable enjoy—it’s just that
which rights these are is going to be determined by justifications that are accessible only
to the reasonable. Still, even when just talking about the distinction between those to
whom justification is owed and those to whom it is not owed it is important to keep in
mind that the latter include many who are, in the natural-language meaning of the term,
perfectly reasonable.
I should note that Quong’s commitment to equal rights to the unreasonable is not
without qualification. Quong is willing to seriously restrict the freedom of speech of the
unreasonable (309). I am not sure, but I think that Quong may be committed to the
justifiability of preventing me from publishing this article, for instance.
29
At least one plausible way of going on the problem of self-defeat has it come out as
a theorem that all non-public-reason-theorists are unreasonable. But perhaps there are
ways of avoiding such a result.
30
The point is not that there’s something strongly counterintuitive about classifying
these people as unreasonable. The point, rather, is first, that the public-reason theorist
can’t have it both ways—if he applies everyday natural-language intuitions regarding
“reasonable” in ways that support his theory (which he shouldn’t do, as “reasonable” here
is a technical term), he must say something about the cases in which they clearly challenge
it; and second, that in the technical sense of “unreasonable”—namely, those justification
to whom is not a necessary condition of legitimacy—excluding all those people is indeed
counterintuitive.
31
Sometimes, public-reason theorists write as if they do. See, for instance, Nagel
(2003, 76); Larmore (2003, throughout).
to reconcile authority with liberty and equality. And we must ask—do such
thoughts apply to the unreasonable as well? The answer seems obvious. The
unreasonable too are born free, and yet they are everywhere in chains.32 And
if by coercing someone based on a principle they do not endorse (nor would
they, after a respectable amount of good reasoning) we fail to treat her as our
equal, then this is so when that someone is unreasonable as well. If you’re
not sure, recall the need to resist the temptation to think of the unreasona-
ble as the Nazis and the murderous psychopaths. Suppose I am quite, well,
reasonable (in the non-technical, natural language sense of this term),
except I reject some of the burdens of judgment. Is there no longer even an
initial tension between my liberty (in whatever sense is important here) and
the authority the state claims over me? Suppose I am motivated to interact
with others as free and equal, but I have other stronger motivations, or
perhaps my conception of what it is to treat others as free and equal is a
little bit off. Is there no longer a problem of equality if you impose on me a
principle I do not endorse? Alternatively, just think of some of the smartest,
nicest, people whom public-reason theorists rule out as unreasonable, and
ask whether the underlying thoughts about liberty and equality apply to
them, at least initially. I can’t see how a negative answer can even get off the
ground. A story reconciling authority with liberty and equality is still very
much needed, with regard to (at least many of ) the unreasonable as with
regard to everyone else.
If the initial problem arises with regard to the unreasonable as well, and
if the restriction to the reasonable is not to be an ad hoc restriction intro-
duced just in order to save the theory, some other rationale must be sup-
plied for the exclusion of the unreasonable. For the most part, this is a need
left unacknowledged by public-reason theorists. Still, at times such ration-
ales are offered (or can be read off what public-reason theorists say). Let me
address the two main ones I find in the literature.33
It is sometimes said that reciprocity considerations are what does the trick
here. The thought seems to be that we should justify ourselves only to those
who, had they had political power, would have bothered to justify them-
selves to us.34 Now, it seems to me an interesting, general question how such
32
I’m paraphrasing Rousseau here, without any exegetical pretensions.
33
It is sometimes hinted that practical considerations—having to do, roughly, with
the costs of including the unreasonable—do the work here. See, for instance, Cohen
(2009, 7), Quong (2011, 37). But I don’t think that these suggestions are in line with the
public-reason spirit—on top of relying on questionable empirical assumptions (seeing
how many people are unreasonable in the relevant sense, excluding the unreasonable may
also be extremely costly), it seems too modus-vivendi-ish for Rawlsians. This is not what
they seem to be after.
34
Macedo (1995, 484), Larmore (2003), Lenman (2010, 179).
124 David Enoch
reciprocity considerations in general behave, and I can’t discuss this ques-

tion in general here (nor do I know what I would say). Let me just quickly
note two points. The first is that reciprocity considerations never seem
to be at the most fundamental moral level. Perhaps, that is, I should only
give you a ride to work if you, had you had a car, would have offered me a
ride. But when it comes to the most fundamental level, this is not so: if
utilitarianism is true, then the persons whose utility we should take into
account are all, not just those who would take our utility into account; and
I don’t recall, in Kant, the proviso that what we should always treat as an
end and not merely as a means is the humanity of only those who would
treat us in a similar way. And we are here in a most fundamental context
indeed—the basis of all political philosophy, the need to legitimize state
authority. Here too, then, reciprocity considerations just seem inappropri-
ate. Those who would “misbehave” politically are still free and equal, and
so subjecting them to authority still calls for justification. Second—to
anticipate the next point—because so many of the constituency of even
current Western democracies are unreasonable, reading the whole project
as based on such reciprocity considerations threatens to divorce it entirely
from the real world.
According to another possible rationale for excluding the unreasonable
the justification for the exclusion of the unreasonable is methodological. We
are doing, at this stage, ideal theory, by which is meant (perhaps roughly) a
theory that assumes full compliance. And it is natural and methodologically
advisable to start off by assuming away some further complications.35 The
point is underdeveloped in Rawls, but perhaps we can fill in some details:
even assuming a political community where everyone is reasonable (in the
relevant sense), still the burdens of judgments apply; so still, reasonable
pluralism is to be expected; and so we can ask—how are we to proceed, in
that society, given such pluralism? The requirement to justify political prin-
ciples (or some such) to each of the reasonable is the answer to that ques-
tion. The unreasonable are irrelevant, because they are not a part of that
society, or an addressee of our ideal theory.
Now, there’s something that seems to me importantly right about this
way of proceeding: namely, the problem is still interesting even when the
unreasonable are assumed away; and philosophers need no further reasons
to engage a philosophically interesting question. Perhaps—though I’m much
less sure of that—it’s even a good idea to accord ideal theory methodological
35
See Rawls’s “The Domain of the Political and Overlapping Consensus” (1989, 236).
That he’s doing ideal theory, and that ideal theory is prior to non-ideal theory, are of
course central themes in Rawls’s A Theory of Justice. But I don’t know of claims in Political
Liberalism where Rawls justifies excluding the unreasonable based on these general
points.
priority (in some sense) over non-ideal theory. But even granting Rawls all
that, still we shouldn’t be tempted by too quick a move from ideal to
non-ideal theory. Even if Rawls is right that some kind of justifiability-to
is a necessary condition for legitimacy in an all-reasonable-community,
it most certainly does not follow that in a community where some are
unreasonable, the analogous condition necessary for legitimacy is
justifiability-to-just-the-reasonable (nor does it follow, if you’ve justified
something to all the reasonable under full compliance, that it’s also justi-
fiable—even just to them—under partial compliance). Unless some
rationale can be offered for this restriction in the non-ideal theory, then
perhaps the thing to conclude is that no justifiability-to requirement is
necessary for legitimacy in non-ideal theory, because none is satisfiable
(because of the presence of the unreasonable). Or perhaps there are some
other ways to go. The general point is simply that it’s never a simple,
immediate step from ideal to non-ideal theory.36 Ignoring this would be
tantamount to trying to design airplanes by relying on mechanical mod-
els that assume frictionless surfaces, just because there’s something to be
said for thinking of some such models for some purposes. So, despite the
intellectual respectability of the ideal-theory way of motivating conduct-
ing a discussion that ignores the unreasonable, this way of proceeding
cannot at the end of the day vindicate excluding the unreasonable, at least
when we’re back to talking (as public-reason theorists seem to) about the
real world.37
36
For similar points in similar contexts, See Cuneo (2013, 360), and Rossi (2014).
This seems to be a point Quong understands well, at least when criticizing others:
“ . . . even if we stipulate that people can only be members of the justificatory constituency
if they accept Gaus’s views on rational justification, this means Gaus’s approach (and by
implication the convergence model) does not apply to the world that we live in, nor does
it apply to a well-ordered liberal society” (272–3).
37
Dreben (2003) is a clear example of someone emphasizing both the ideal-theory
nature of the project (e.g., 323, 344) and the thought that important conclusions about
the real world follow (e.g., 317, 328). He doesn’t do anything to reconcile these arguably
inconsistent claims.
I think that Quong’s recent (2011) emphasis on the internal nature of the political
liberal project is in line with the Rawlsian ideal-theory way of excluding the unreason-
able (in conversation, Quong has confirmed this), and at times (e.g., 143) he even talks
explicitly in terms of ideal theory. Now, I think that Quong’s insistence on the internal
project cannot withstand criticism in general, but he develops this line in great detail,
and offers explicit justification for doing it. I remain unconvinced, and I hope to say
why in future work. Here, I must settle for noting that as a response to the problem of
excluding the unreasonable, Quong seems to fail for reasons similar to the ones in the
text here. See especially 290, where he seems to be guilty of precisely the fallacious
move from ideal to non-ideal theory highlighted in the text—having justified some-
thing under full compliance, proceeding to assume that it’s been justified to the com-
plying under partial compliance.
126 David Enoch
Notice that my point here—perhaps unlike some other current discussions

of ideal and non-ideal theory in political philosophy38—strikes at the very
heart of public-reason accounts: their suggested necessary condition for legit-
imacy. The point is not just about the unsoundness of drawing more specific
practical implications from an ideal theory to the real world (though it is that
too); nor is my point about the interest in doing ideal theory (of course, that
too is interesting and valuable—many things are); nor is the point about the
importance of aiming at the truth regarding political matters, even when
doing so will have no effect in the world (I agree that this too is of importance,
and a part of the political philosopher’s job). The point, rather, is that if the
only way for the public-reason theorist to motivate excluding the unreasona-
ble is by resorting to ideal theory, she can no longer maintain that justifiability
to all the reasonable (but not to the unreasonable) is a necessary condition for
legitimacy in the real world, where not everyone is reasonable. All she has to
say, then, is that justifiability to all would be a necessary condition for legit-
imacy in an all-reasonable environment. I doubt that many public-reason
theorists will be happy with this result,39 and I want to insist that they
should not. (I’ll have more to say about how going down this road betrays
the role of the political philosopher later on.)
The underlying problem of reconciling the state’s authority with the free-
dom and equality of those subject to it arises with regard to the unreasona-
ble just as it does with regard to the reasonable, and no rationale that I can
think of or that I find in the literature can motivate the exclusion of the
unreasonable from the constituency of the public-reason requirement con-
sistently with the underlying public-reason motivations. So, excluding the
unreasonable in this way is objectionably ad hoc—it is a revision of a the-
ory, introduced in a way that’s inconsistent with its underlying motivations,
just in order to save it from counterexamples. So, the thing to do given the
obvious counterexamples is not to exclude the unreasonable, but to reject a
justifiability-to requirement.
3.4 Hypothetical Conditions

So much, then, for the first mechanism employed by public-reason theorists
to avoid the problem that if we take people as we find them, nothing is
38
See, for instance, Farrelly (2007), Stemplowska (2008), and Simmons (2010).
Unlike Simmons, I am not invested in any interpretive claim about how best to under-
stand these terms in Rawls. And notice also that the point I am about to make in the text
stands even if everything Simmons says (including about the priority of ideal theory, in
the sense he finds in Rawls) is true.
39
Quong seems to be officially okay with this result, but—as can be seen from the
references in footnote 37 above—doesn’t seem consistent in this regard.
justifiable to all of them (on any plausible understanding of justifiable-to).

The second mechanism mentioned above is that of going hypothetical. What
is needed for legitimacy, the thought is, is not necessarily that the relevant
principles (or whatever) be justified to the satisfaction of members of the
relevant constituency as we find them, but rather that they would be satisfied,
if placed in the right hypothetical conditions. Such conditions may include
things like being fully informed, perhaps satisfying some norms of proce-
dural rationality, thinking about things in a calm moment, and so on.
But now we should ask of this idealization, or hypotheticalization,
whether it passes the test from section 3.2 above. Is it well motivated, in
a way that is consistent with the motivations underlying public-reason
accounts in the first place? Or is it an ad hoc “fix,” introduced merely in
order to avoid some obvious counterexamples and save the theory?
Think again of consent and related normative phenomena. Perhaps I owe
you some money, because of a contract I signed. You come asking for your
money, and I ask why I should give it to you (myself, I would rather use
it in different ways). You note that I consented, or promised, or signed a
contract, or some such. This seems, at least sometimes, as a good response.
But now suppose there never was any contract. Still, you are not at a
loss for words. To my question (why should I give you the money) you
respond: “I know you didn’t give your consent. But you would have, if
you spent a respectable amount of good reasoning on it,” or utilizing some
other hyotheticalization of this kind. This sounds like no answer at all.
In this case, hypothetical consent is no more substitute to consent than
hypothetical water is to water for the thirsty. Of course, you may want to
indicate that I should have consented, that I had overwhelming reasons
to. But if you do that, it’s no longer clear what work is being done by the
hypothetical consent—you can explain why I have to pay directly by
appeal to these underlying reasons. This is a common, well-known prob-
lem for hypothetical consent theories.40 And a similar initial worry seems
warranted in the context of a justification-to principle as well. It’s not
clear how the fact that a justification would have satisfied me had I been in
some hypothetical conditions goes any way at all towards reconciling
authority and my liberty, given that I am not actually satisfied with the
justification offered. True, perhaps I should be so satisfied. But then, pre-
sumably there are reasons why I should be satisfied with the justification
offered, and then all the work is being done by those reasons directly, not
40
Again, see Dworkin (1973). For a recent attempt to respond to it—though in the
context of his specific, somewhat idiosyncratic, normative hypothetical consent theory—
see Estlund’s (2008) discussion of normative consent.
128 David Enoch
by their availability to me. The justification-to requirement has been left

out of the normative picture.41
If this were so—if no hypothetical satisfaction with an offered justification
ever did any normative work—then public-reason theorists would have to
choose whether to go for actual satisfaction (and embrace the anarchism that
would follow) or to concede the failure of the justification-to requirement
and with it of the public-reason tradition in general. But I think that this
would be too quick. For sometimes, even in discussing consent, hypothetical
consent does matter. Perhaps, for instance, mild, superficial corrections of
procedural irrationality are okay in this respect—perhaps, in other words, if
you don’t consent to a medical procedure because you’re too drunk to listen
to the explanations you yourself would have accepted unhesitatingly a couple
of drinks ago, then for some purposes the situation is similar to one in which
you did give your consent.42 So, hypothetical consent—and hypothetical
justifiability—are not just like hypothetical water. Sometimes they make a
normative difference.
Ok then, but when? Under what conditions does the fact that one—
dissatisfied as one actually is with the offered justification—would be justi-
fied in some suitably specified hypothetical conditions make the needed
normative difference? I don’t have a general answer to this question (I hope
to have more to say on it in the future). And so the discussion here will not
be complete. But it seems to me that even in the absence of such a complete
discussion, we can return to the circumstances of political disagreement and
argue that hypotheticalizing here is of no help at all.
The way to see this is to think about a real-life political disagreement. The
public-reason theorist claims the status of legitimate authority for some prin
ciple on which political action is to be grounded—say, one allowing for a
rather extensively redistributive scheme of taxation; a (reasonable) citizen
comes along, rejecting the justification offered for the principle or policy;43
furthermore, she challenges the legitimacy of the principle, because it has
not been justified to her, to her satisfaction; the public-reason theorist is
undeterred: for legitimacy what is required, he reminds her, is not that the
principle be justified to her satisfaction, but that it be justifiable to her sat-
isfaction, so that she would be justified, in the suitably specified hypotheti-
cal conditions. Perhaps, he insists, she hasn’t been paying enough attention
to the arguments, and would have been convinced had she done better in
this regard; or perhaps she’s being in some other way procedurally irrational;
41
“There may be no middle way between actual (including implied) agreement and
rational justification” (Raz 1990, 46).
42
This often-used kind of example comes, for instance, from Waldron (1987, 132).
43
See, for instance, Gaus (2011, Chapter 24).
perhaps she’s not well informed; perhaps, while satisfying the minimal rea-
sonableness criteria, still she hasn’t been taking others’ point of view into
account sufficiently impartially. Because this is so, the public-reason theorist
reassures himself, there is really no legitimacy problem after all. All is well.
How convincing is this line of thought, as applied to real-life political
disagreements? Remember, we’re not talking about the highly anxious or
drunk patient, in the emergency room, in an especially vulnerable time in
his life. We’re talking about serious people taking part in serious discussions
over long periods of time, in the public sphere or in the privacy of smaller
social interactions, and anyway very deeply committed to their respectable
views. And let’s set aside for now other related worries, such as how civil it
would be to say to the one rejecting the principle that he would have
accepted it had he spent a respectable amount of good reasoning on it.44 The
question we’re interested in is whether such a response is in line with the
underlying public-reason motivations. If I am the person rejecting the sug-
gested taxation scheme, I am in the Rousseauean predicament—supposedly
free, but about to be coerced to take part in this policy which I reject. Is this
tension in any way reconciled seeing that I would have accepted the princi-
ple in some hypothetical conditions that are sufficiently far so that I am not
moved here and now, not even having thought about things quite seriously,
having debated them with my friends, and so on?
I am inclined, of course, to say that this is no response at all. The tension
between authority and liberty is fully present in these cases, even given this
kind of hypothetical justification-to. Now, let me emphasize again that I am
not claiming that hypothetical consent and justification-to never matter.
What I claim is that the thought that they matter here, where they are sup-
posed to matter most, in the context of political disagreement, resists belief.
Also, note that all of this is supposed to be from the point of view of a
public-reason theorist, one who thinks that justification-to is extremely
important. My point is not that it’s extremely important to justify the taxa-
tion scheme to all those rejecting it. My point is just that if you think that
justification-to is a necessary condition for legitimacy, because this is the
only way to reconcile authority with liberty and equality, or to respect as
equal and free your fellow citizens, then you should not be impressed with
the response at the end of the previous paragraph. That response amounts
to no progress in respecting your fellow citizens as free and equal. It should
not impress you more, for instance, than the response “I understand that
you are not convinced, but I’ve already stated my reasons for why it’s true
44
For this point, put in terms of adding insult to injury, and directed at Gaus, see my
(2013, section 6). Hampton (1993, 309) seems to anticipate something like this line. And
Wolterstorff (2012, 74) makes precisely this point as well.
130 David Enoch
that this taxation scheme is justified.”45 And this response is, of course, not
one that a public-reason theorist can settle for without abandoning his com-
mitment to public reason.
Furthermore, the public-reason theorist who wants to use such idealiza-
tion owes us also the details—what the idealization does and what it does
not rule out. It’s not going to be easy to fill in the details in an extensionally
adequate way. And, what’s more, extensional adequacy is not enough. It’s
also necessary to offer a rationale for the specific way of idealizing, or the
specific hypothetical conditions deemed relevant, a rationale that’s consist-
ent with the underlying public-reason motivations. I don’t see how this can
possibly be done.
Going hypothetical, I conclude, is no more promising a strategy for the
public-reason theorist than excluding the unreasonable. Any such attempt
will be objectionably ad hoc. The public-reason theorist should either settle
for actual justification-to as a necessary condition for legitimacy (which in
all likelihood entails anarchism, even when restricted to just the reasonable),
or else take back his commitment to public reason.
4. THE REASONS WE ACT ON (FOR INSTANCE,

WHEN COERCING)
Recall the thought that by refusing to go sufficiently neutral—by relying on

claims that are not justifiable to the entire relevant constituency—we fail to
treat those to whom it is not justifiable as equal. The thought seems to be
that if I am willing to impose Catholicism-based imperatives on you,
non-Catholic as you are, I am giving extra political weight to my own beliefs
over yours. I am treating in an asymmetric way the fact that I believe
Catholicism and the fact that you believe its denial. According equal weight
to your beliefs and mine would require going neutral here.
This thought, though, is based on a confusion about reasons for actions.46
When I impose (in good faith) Catholicism-based directives on you, my
45
A similar point comes up in the context of granting conscientious exemption from
military service. Statman (2009) argues that in the Israeli context pacifist objectors are
wrong about what follows from their own pacifist commitments, so that refusing to
exempt them is no attack on their conscience. I (2009b) respond by noting the point in
the text: The tension between freedom of conscience and the demands of the state is not
in any way relieved by the fact (if it is a fact) that the objectors are wrong about what
follows from their own deepest commitments.
46
This confusion pervades, I believe, Rawls’s and Rawlsian texts. For one clear exam-
ple, see Political Liberalism (1993, 129). Usually, though, the confusion is less explicit than
that. Still, attributing this confusion to Rawlsians is needed in order to make sense of
claims that equality is in any way relevant here.
reason for action—that is, the feature of the circumstances that I take to
be normatively relevant—is not that I believe Catholic doctrine. Rather,
my reason is the content of that belief, namely Catholic doctrine itself.
Similarly, if I, a comprehensive, “metaphysical” liberal impose directives that
are based on the value of autonomy on people who do not value autonomy,
my reason for action is not that I value autonomy. Had this been the reason,
then given that they do not value autonomy, privileging my valuings over
theirs would have violated some intuitive principle of equality. But my reason
for action is different, and is not about me at all. Rather, my reason for action
is that autonomy makes people’s lives go better, or some such. There is nothing
indexical about this reason for action, and so no violation of equality at all.
The point is tricky, and so merits more elaboration. It’s tricky partly
because it’s hard to distinguish, from the first-person perspective, between
that-p and that-I-believe-that-p. But with the help of some counterfactuals,
this can be done. We’re assuming, then, that in the actual world, you think
that autonomy is of value. Let’s imagine a possible world in which you are
mistaken about autonomy, thinking that it’s not of value, even though it is.
Do you still want—in the actual world—to impose autonomy-based direc-
tives on people in that hypothetical world? A positive answer seems obvious.
But in that world, you don’t believe in the value of autonomy. Still, in that
world autonomy is of value. So what we can learn from the fact that when
you think (in the actual world) about that possible world you still want to
impose autonomy-based directives is that your reason for imposing them is
that autonomy is of value (which is true there too), not that you believe that it
is of value (because on that world, you don’t). Compare this to cases of, say,
taste. Suppose that in the actual world you prefer vanilla ice cream to choc-
olate. Now imagine a possible world in which you prefer chocolate. And
suppose that you—in the actual world—have to order ice cream for your
counterpart in that possible world. What flavor do you order? Clearly, choc-
olate. What this shows is that in the actual world too, when you order
vanilla ice cream, your reason for action is indexical, it is about you and your
preferences, it’s simply that you prefer vanilla. This is why it makes no sense
to order vanilla ice cream for your counterpart in a world in which he no
longer prefers vanilla. But when you act on a principle or something like the
value of autonomy, this simply isn’t so. Your reason for action in such cases
is that-p, not that-you-believe-that-p. It’s just not about you at all. So, there
is no sense in which you’re giving extra weight to your beliefs over others’—
you’re giving no weight to your beliefs here. You’re just giving weight to the
value of autonomy, or the truth of Catholicism, or some such.
This point is in no way special to the political realm. It applies across a
wide range of cases, in practical as well as theoretical reasoning, regarding
justification and explanation alike. Suppose you believe that what explains
132 David Enoch
the tide is the gravitational force of the moon. What, do you believe, does
the explanatory (and causal) work here—that the moon has this-and-that
gravitational force, or that you believe that it does? Clearly, the former. You
don’t believe that your beliefs causally regulate the tides, nor do you think
they explain them. What has relevant explanatory force is the content of
your beliefs, not the fact that you have them. We can run the counterfactual
test again—surely, when you think (in the actual world) about a possible
world in which you’re wrong in your beliefs about the explanation of the
tides, you don’t (in the actual world) think that the explanation of the tides
in that possible world is any different than the one in the actual world.
What does the explanatory work is entirely present there—it is the moon’s
gravitational force, not your beliefs about it (which are absent there). The
explanation, then, is not about you at all.
Similarly, in epistemology the question sometimes arises whether—in a
case of disagreement with someone you take to be your epistemic peer on
the relevant question—you should give extra weight to your own evalua-
tion of the relevant evidence. Here too, though, it’s clear that in typical
cases, the thing to which you accord epistemic weight is not that you eval-
uate the evidence a certain way, but rather the evidence and its epistemic
value. You are not a part of the picture of what does the epistemic work. It’s
just not about you.47
And the same point applies to our reasons for action even when they are
not themselves normative. If I believe that cutting the blue wire will defuse
the bomb, you believe that cutting the red one will, and I proceed to cut the
blue one, there is no sense in which I am giving extra weight to myself or my
beliefs. My reason for action is not that I believe that cutting the blue wire will
defuse the bomb. This is not the feature of the case to which I accord norma-
tive significance (as can be seen, for instance, by running the counterfactual
test again). Rather, my reason for cutting the blue wire is that doing so will
defuse the bomb. And this reason is not indexical at all. Once again, then, it’s
not about me.48
47
This is one of my main points against the Equal Weight View in “Not Just a
Truthometer” (2011b). For similar points in a wider epistemic context, see Schroeder
(2008). Of course, not many things are uncontroversial in the peer disagreement litera-
ture. But the point in the text, it seems to me, should be. (Those I criticize in “Not Just a
Truthometer” do not argue against this point; rather, they seem to either assume its denial,
or not to see its full significance.)
48
You may be worried about mistakes: what if in fact autonomy is not of value? Surely,
in that case my reason for action can’t be that autonomy is of value (for it isn’t), and the only
remaining candidate seems to be my belief in the value of autonomy. But this is not so. A
fuller discussion of these issues—like the one I offer in (2011a, 221–3; see also the references
there)—will distinguish between different senses of “reason for action,” only some of which
are factive. In the terms I use there, the agent’s reason—the feature of the situation the agent
Now, you may think that still there is a sense in which I am not out of the
picture when it’s the content of my beliefs that plays the relevant role in my
reasons for belief or action. And I think that this is right, but we have to be
careful about the precise nature of this role.49 The point is, again, that the
fact that it is my view is not a part of what is, on my view, normatively rel-
evant (unlike, for instance, in the mere preference for vanilla case). It is thus
not a part of my reason for action. It may be a necessary background condi-
tion for my reason being a reason for me to act in the relevant way. But this
is very different:50 something about me may be a part of the story of why I
offer the value of autonomy as a reason for (political) action. But it’s not a
part of the reason itself. Again, the situation in other cases is exactly alike:
perhaps the fact that I believe that the moon’s gravitational force explains
the tides is a part of the background conditions needed for my offering that
explanation of the tides. But this belief of mine is not a part of the explana-
tion itself. The explanation is not about me.
Still, you may want to insist, even if nothing about me is a part of my
reason for action, still I am involved in this other, background-ish kind
of way. Furthermore, even if I do not, by acting politically on the value of
autonomy, give extra weight to my belief about the value of autonomy
(because that belief is no part of my reason for action), still I act in a way that
will, as I know, bring about a situation in which my beliefs are more in line
with the state’s actions than the beliefs of those rejecting the value of auton-
omy. In this sense too, then, it may still be argued that acting on the value of
autonomy, in the face of (reasonable?) disagreement about it, violates equal-
ity. But this would be a mistake. It’s true that when I act, I act on my own
beliefs. There is no way around it: anything I do is something that I do. Even
if I decide to defer to another, it’s me doing the deferring, and indeed the
deciding to defer. Even if I become a card-carrying member of the public-
reason club, the refusal to act politically on my comprehensive doctrine will
be very much my refusal. My beliefs and actions are and always will be my
beliefs and actions.51 Thus, if by acting on the value of autonomy I violate
equality (because it’s my own beliefs I act on), then any action whatsoever
violates equality. This is not the kind of equality worth caring about.
finds normatively relevant—need not be factive. Even in the case in which autonomy is not
of value, then, my reason for action may be the value of autonomy (just like even if the
moon is not after all responsible for the tides, still what in my view does the explanatory
work is the moon’s gravitational force, not my belief about it).
49
Raz (1998, 27) puts this point correctly but somewhat cryptically thus: “We rely on
the answer to the question not because it is our view, but because it is, as we believe, true.”
50
This distinction is central to Mark Schroeder’s work. See, for instance, his (2007,
Chapter 2).
51
See Raz (1998, 27).
134 David Enoch
What the discussion in this section shows is that the equality underlying
public-reason motivation is based on a simple confusion. The liberty motiva-
tion should be, to an extent, accommodated (a point I get to below). But the
equality motivation should just be rejected. The tension between authority
and equality (as understood by public-reason theorists) is a pseudo-problem.
5. THE ROLE OF THE POLITICAL PHILOSOPHER
The Rawls of Political Liberalism is sometimes accused of blurring the lines

between political philosophy and politics:52 the emphasis on stability that
seems to enter the picture much earlier than it would for most others, the
danger that his account is “political in the wrong way,”53 distancing oneself
from truth and replacing it by a standard that seems to depend at least
partly on what the people around us accept, the attempt to apply a principle
of toleration to philosophy itself 54—all these make the suspicion that Rawls
has become a bit of a politician understandable. And Rawls and Rawlsians
are, of course, eager to show that this isn’t quite so. What I want to do in
this section, though, pulls in the opposite direction. For I want to high-
light—regardless of the details of the controversy just noted—a way in
which the public-reason tradition creates too big a divide between political
philosophy and politics.
If you’re a public-reason theorist, how do you view those fighting in the
political arena for everything they think is good and just? Perhaps some of
them, of course, are themselves committed to public reason, so that they’re
really fighting only for what’s reasonable (or some such). But much more often,
the rhetoric and also sincere beliefs of political actors and activists is that of
comprehensive doctrines. They have a worldview, and one that is much too
rich and exciting to be acceptable according to a public-reason justifiability-to
requirement. With regard to these, the public-reason theorist—in his capacity
as a political philosopher—refuses to engage. Of course, he too may have his
“private” moments, in which he engages. But when he does that, he must
think of himself as “no longer engaged in political philosophy suitable for a
well-ordered liberal society . . . ” (Quong 2011, 242). In his capacity as polit-
ical philosopher, he transcends the political arena, and rather than engage
the disagreements within it, merely views it as the data to be taken into
account, perhaps as trouble to be contained.
52
See, for instance, Raz (1990, 10).
53
See Hampton (1989). Rawls’s (1989) is motivated in large part by the need to
respond to this accusation.
54
Rawls (1993, 10).
There is a huge difference between thinking about a disagreement “from

within,” and merely taking the fact of disagreement as something to be
handled, “from the outside.” Think of any politically controversial issue
about which you have a view, perhaps one you care about rather passion-
ately, where you’re rather confident that you are right: say, just in order to
have a concrete example, whether the USA should intervene in the situ-
ation in Syria. And think about what it is like to engage those on the
other side of this controversy. Now think about what it is like to engage
in discussion (or deliberation) not over the matter at controversy itself
(whether to intervene in Syria), but over how to deal with the fact that
this question is controversial. This latter question too, of course, can be
politically important and theoretically interesting. But crucially, it’s a
different question.55 And just as crucially, the fact that this second ques-
tion is there doesn’t mean it’s the only important question there. Surely,
the first question—the first-order question about which people disagree—
is also an important one to ask. But the public-reason political philoso-
pher insists that so long as it’s one about which reasonable citizens may
differ, it’s one he—and political philosophy in general—has nothing to
say about.
What is involved in thinking about the disagreement in this way, in
refusing to engage the parties to the disagreement on the merits of the topic
they’re disagreeing about, merely thinking about how to contain the disa-
greement’s danger and harm?
Clearly, this involves thinking about the disagreement in a way that’s very
different from the way in which the parties to the disagreement themselves
are thinking about it. For them, the crucial question is what the USA should
do in response to the crisis in Syria, not what should be done about the dis-
agreement about how to proceed vis-à-vis that crisis. Furthermore, it’s hard to
resist the feeling that there’s something patronizing about the public-reason
attitude to the disagreement: it’s as if the public-reason theorist thinks of
himself as the responsible adult, viewing the children quarreling, refusing to
engage the details of the quarrel, just trying to see to it that no serious harm
55
There may be more questions still. One (i) is the first order question, on its own.
Another question (ii) is the one in the text—about how to deal with disagreement.
Possibly there’s a third one—revisiting the first order question, now with the answer
to the second one in mind; roughly this is the question (iii) whether to intervene in
Syria given the disagreement, or taking into account also the fact that people differ
about (i). One doesn’t have to be a public-reason theorist to acknowledge a difference
between (i) and (iii). Disagreement may be one normatively relevant fact here, among
others. (And once we see this, we may also ask many more questions: (iv) how to
respond to (iii) given that there’s disagreement about it as well; (v) how to answer (i)
given disagreement about (i) and (iii); (vi) how to answer (v) given disagreement
about it; . . .).
136 David Enoch
is caused as a result of the quarrel. Respecting those engaged in the debate—

indeed, respecting them as free and equal responsible adults—would here
amount to engaging them, not to retreating from the arena, “attempting to
occupy a noncontroversial high ground.”56
By refusing to engage in this way, the public-reason political philosopher
treats some of the most deeply held beliefs of those engaged in the relevant
disagreement as if they were mere preferences: for with mere preferences,
arguably going second-order and impartial is precisely the way to go, and
certainly engaging them in argument doesn’t make much sense. Indeed,
returning now to the lesson of the previous section, with mere preferences
the normatively relevant fact is that people have them, not anything about
their content directly—the reason for action of someone acting on her pref-
erence for chocolate over vanilla is the indexical one “that I prefer choco-
late.” But the reason for action of someone acting in the public arena on his
belief that the USA should intervene in Syria is not indexical in this way at
all—it’s not about his preference for intervention, but about intervention
being (as he thinks) the right way to proceed57 (or about the normatively
relevant circumstances in virtue of which it is right). The public-reason the-
orist misses all this, and relates to such principled disagreements as if they
were merely preference-based.58 And because this is most certainly not how
the parties to the disagreement think of their commitments and disagree-
ment, the public-reason theorist fails to respect them and their commit-
ments. Hence, the patronizing.
This means that, appearances to the contrary notwithstanding, there’s a
sense in which public-reason theorists, with their emphasis on reasonable
pluralism as a fact any acceptable political philosophy should take into
56
Not just this quote comes from Raz (1998, 47), but also the thought expressed here
more widely: it echoes Raz’s insistence that treating people with respect requires engaging
them, not, for instance, propagating false beliefs (1998, 43). I think that Raz overstates
his case—respecting people amounts to different things in different contexts, and per-
haps sometimes is best manifested by a refusal to engage them. Not, though, in the
context in the text, certainly not in the patronizing public-reason kind of way.
57
For a close point see Raz (1998, 35, footnote 13).
58
Nagel (1987, 158) does notice this, as when he says: “Impartiality among persons is
one thing, but impartiality among conceptions of the good is quite another.” But at the
end of the day he thinks that this gap can somehow be bridged. In another (metaethical)
context, I discuss in detail the distinction between conflicts that are based on mere pref-
erences and those that are based on moral disagreement. See Chapter 2 of my Taking
Morality Seriously (2011). There is some similarity between the point in the text and an
oft-made criticism of political liberalism, according to which it prevents (for instance) the
religious from relying on their deeply held beliefs in the public domain. But the point in
the text is different—it’s not about what considerations political actors are allowed to cite
in the public domain, exactly, but about how the public-reason theorist thinks of his own
role vis-à-vis such debates.
account, do not take disagreement seriously enough. Taking the disagree-

ment over intervention in Syria seriously means being willing to engage it,
not retreating to a patronizing neutral high ground.59 Of course, even while
engaging the substance of the first-order disagreement, second-order ques-
tions may be asked. And as even the non-public-reason political philoso-
pher will agree, there are restrictions on how to engage the first-order
disagreements. The fight for everything that is just and good is itself some-
times constrained by substantive considerations. Political philosophy, in
other words, is hard. Surely, this should come as no surprise, to political
philosophers and activists alike.
You may think that the debate has gone somewhat verbal, with public-
reason theorists and others talking past each other, as if simply different things
are meant by the two parties when they employ such terms as “political phi-
losophy” (this could explain also the they-just-don’t-get-it feeling on both
sides). And I’m sure that that’s a part of what’s going on, but it can’t be the
whole thing. It’s not a controversy that can be eliminated by disambiguation.
There is a different vision involved, and one that itself should be the subject of
substantive discussion: for it seems to me that the good political philosopher
and the political activist—at the very least, the activist who is fighting on the
side of the right and the good—should be allies. Indeed, their roles are in a
way continuous. True, there are differences—perhaps, for instance, there are
small-scale pragmatic considerations that the political philosopher can ignore
and the activist should not. Or perhaps the activist may be excused for
argumentative imprecisions that cannot be tolerated from the political phil
osopher. And certainly, the political philosopher should (sometimes) engage
matters on a level of abstraction that is unsuitable for activist texts.
Indeed, the political philosopher may from time to time engage in find-
ing interesting truths that have no practical bearing whatsoever—to repeat
a point from section 3.3 above, that they’re interesting should be reason
enough for philosophical inquiry. But these differences between the politi-
cal philosopher and the activist notwithstanding, the activist and the political
philosopher are engaged in what is largely the same project. The public-
reason vision for political philosophy is different, though. On it, the political
philosopher looks at activists from a distance, and with a patronizing eye.
On this view, it’s not just that not all political philosophy is guaranteed to
be relevant for activism; rather, it’s that virtually all political philosophy is
guaranteed not to be relevant to activism. This is not what political philoso-
phy should be like.60
I make a somewhat similar point against Waldron in my (2007).

59
Nor, it seems to me, is it what it has been like in the field’s long history (most of
60
which, I confess, I do not know).

138 David Enoch
6. ACCOMMODATING (TO AN EXTENT) THE

UNDERLYING PUBLIC-REASON INTUITIONS
If I am right, then, there is no hope for a recognizably public-reason view.

There is no remotely plausible way of offering a necessary condition for
legitimacy in terms of justifiability to the relevant constituency, in a way
that remains relevant to the underlying motivations of reconciling authority
with liberty and equality. But what of the intuitions that led us down this
path in the first place? Think again about Estlund’s Pope. Even if the thought
that by subjecting the non-Catholic to the Pope’s directives we’re failing to
treat her as an equal has been conclusively debunked (in section 5), still
there’s something problematic in so doing, isn’t there? Can this fairly robust
intuition be accommodated, without going all the public-reason way?
I think it can. But we should be careful about what exactly the raw, robust
intuition is. The intuition is not exactly about necessary conditions for legit-
imacy, or about the need to justify-to, or any such thing: these are much too
theoretical matters, and they come into play later, as it were, not at the level
of initial intuitions. The raw intuition is the one I just stated—namely, that
even if Catholicism is right, and the Pope infallible, still there is something
objectionable about subjecting a non-Catholic to the Pope’s directives. It is this
powerful intuition that needs to be accommodated.
And it rather easily can.61 For very plausibly, it’s a good thing to engage
people (roughly speaking) on terms that they can accept, pretty much as
they actually are. Perhaps they are even pro-tanto owed to be engaged in
such a way. When other things are equal, a political arrangement is better to
the extent that it manages to justify itself to the relevant constituency, to the
extent that its justification is accessible to its subjects pretty much given
their deeply held beliefs, principles, reasons. We can give a deeper story for
why this is so—I would give one in terms of autonomy, but perhaps there
are other possible stories. The crucial thing is that for whatever deep reason,
there’s something to be said for justification-to, or something similar to it.
Note that on this suggestion nothing about justification-to is a necessary
condition for legitimacy. Rather, it’s yet another political desideratum, yet
another item on the list of significant factors counting in favor of a political
arrangement. It has no lexical priority over others, and is to be balanced
against them. This suffices to fully explain Estlund’s Pope example: there is
61
I don’t know of anyone pursuing this line, but there are some related hints in the
literature. The one most clearly flirting with this line is Brower (1994, 22). And there
seems to be something in this direction also in Hampton (1989, 801), Nagel (2003, 77),
Estlund (2008, 163), and Kelly Sorensen’s (2013) “aspirational political liberalism.”
indeed something objectionable about subjecting the non-Catholic to the

Pope’s directives, for a political order that does that is less good for doing so.
Similarly, for a political order that imposes autonomy-based directives on
the members of an authoritarian illiberal community. But this doesn’t mean
such imposition is always and everywhere illegitimate—whether it is depends
on whether it can gain in other values (or, in the latter case, in the same
value, the value of autonomy) what it loses in failing a plausible justifica-
tion-to requirement.
The advantages of this way of accommodating the underlying public-
reason intuition should at this point be clear. Because justifiability-to is no
longer considered a necessary condition for legitimacy, its failure with
regard to some does not entail anarchism. This means that there is no need
to exclude the unreasonable, or to idealize. On the other hand, depending
on the underlying, perhaps autonomy-based, story, there may be room for
some gradability. Perhaps, for instance, while it’s always pro-tanto objection-
able (other things being equal) to fail to justify the political authority to
someone subject to it, still it’s worse when that person is reasonable (in the
natural-language sense of this term) than when she is not. Or perhaps a
substantive understanding of the value of autonomy entails that choices
under some mildly hypotheticalized conditions are more important for
autonomy than those made under unfavorable actual conditions. If so, we
get some plausible, gradable, well-motivated ways of giving weight to some
of the more plausible thoughts in the vicinity here (perhaps about false
consciousness, or about the drunken patient). But we do so without render-
ing the account vulnerable to the difficulties from section 3 above.62
And because no claim is being made about some spurious neutrality,
and because the political philosophy that emerges from this picture is one
that is avowedly based on evaluative and perhaps other normative judg-
ments, this way of thinking of political philosophy is entirely consistent
with the vision described in the previous section: the political philosopher,
on this picture, acknowledges that all there is is the arena itself, and enters
it, armed with the list of desiderata on which justifiability-to is another
item, to fight, shoulder to shoulder with non-philosophers, for everything
that is just and good.
I realize, of course, that many public-reason theorists will not be willing to
settle for this more minor role of accessibility or availability or justifiability-to.
And in all honesty, the value of justifiability-to will often be outweighed
by others, as the list of political desiderata is long and dense, and the stakes
62
Similarly for the self-defeat worry (again, see Wall (2002), and my (2013)). And I
am pretty sure that this is true also of the epistemological problems facing public-reason
accounts, but I have to fill in all the details there.
140 David Enoch
in politics usually high. Still, this way of accommodating the underlying

public-reason intuitions is the most that is directly supported by those intu-
itions, and more importantly, the most that can withstand criticism.
Let me not pretend that our work is done. Many more details are needed:
we need an account of why it is that justifiability-to is of value; we need to
understand what are the features that determine of how much value it is in
specific circumstances; the relevant kind of access (or justifiability-to) has to
be described, understood, and motivated; and so on. But the discussion
above shows, I think, that this is the work that needs to be done.
Nothing at all about justifiability-to is a necessary condition for legiti-
macy. Still, justifying political arrangements to those subject to them (in
some, yet to be made precise, sense of “justifying to”) is of pro-tanto signifi-
cance. When other things are equal, it should be strived for. This is all we
can get, and all those tempted by public-reason theory should have ever
gone for.
Bibliography
Brower, Bruce (1994). “The Limits of Public Reason,” The Journal of Philosophy 91,
5–26.
Burton, Dreben (2003). “On Rawls and Political Liberalism,” in Freeman, 316–46.
Cohen, Joshua (2009). “Truth and Public Reason,” Philosophy and Public Affairs 37,
2–42.
Cuneo, Terence (2013). Book Review of Paul Weithman’s Why Political Liberalism?,
Faith and Philosophy 30, 357–61.
Dworkin, Ronald (1973). “Justice and Rights,” in his Taking Rights Seriously
(Cambridge: Harvard University Press), Ch. 6.
Enoch, David (2005). “Why Idealize?,” Ethics 115(4), 759–87.
Enoch, David (2007). “Taking Disagreement Seriously: Some Critical Comments
on Jeremy Waldron’s Law and Disagreement,” The Israel Law Review 39, 22–35.
Enoch, David (2009a). “On Estlund’s Democratic Authority,” Iyyun 58, 35–48.
Enoch, David (2009b). “More on the Conscience-based Exemption from Military
Service: A Reply to Dani Statman,” 31 Iyunei Mishpat (Tel Aviv University Law
Review), 709–40 (in Hebrew).
Enoch, David (2011a). Taking Morality Seriously (Oxford: Oxford University Press).
Enoch, David (2011b). “Not Just a Truthometer: Taking Oneself Seriously (But Not
Too Seriously) in Cases of Peer Disagreement,” Mind 119, 953–97.
Enoch, David (2013). “The Disorder of Public Reason: A Critical Study of Gerlad
Gaus’s The Order of Public Reason,” Ethics 124, 1–36.
Enoch, David (2014) “Authority and Reason-giving,” Philosophy and Phenomeno
logical Research 89, 296–332.
Enoch, David (Manuscript). “Idealizing Still Not Off the Hook,” available at
<http://law.huji.ac.il/upload/ReplyToSobel.doc>.
Estlund, David M. (2008). Democratic Authority: A Philosophical Framework

(Princeton: Princeton University Press).
Estlund, David M. (2009). “Reply to Commentators,” Iyyun 58, 73–88.
Farrelly, Colin (2007). “Justice in Ideal Theory: A Refutation,” Political Studies 55,
844–64.
Freeman, Samuel R. (ed.) (2003). The Cambridge Companion to Rawls (Cambridge:
Cambridge University Press).
Gaus, Gerald (2011). The Order of Public Reason (Cambridge: Cambridge University
Press).
Hampton, Jean (1989). “Should Political Philosophy Be Done without Metaphysics,”
Ethics 99, 791–814.
Hampton, Jean (1993). “The Moral Commitments of Liberalism,” in David Copp
and Jean Hampton (eds), The Idea of Democracy, 292–314.
Johnston, Mark (1989). “Dispositional Theories of Value,” Proceedings of the
Aristotelian Society 63, suppl., 139–74.
Larmore, Charles (1990). “Political Liberalism,” Political Theory 18, 339.
Larmore, Charles (2003). “Public Reason,” in Freeman, 368–93.
Lenman, James (2010). “Humean Constructivism in Moral Theory,” Oxford Studies
in Metaethics 5, 175–93.
Macedo, Stephen (1995). “Liberal Civic Education and Religious Fundamentalism:
The Case of God v. John Rawls?,” Ethics 105, 468–96.
Nagel, Thomas (1987). “Moral Conflict and Political Legitimacy,” Philosophy and
Public Affairs 26, 215–40.
Nagel, Thomas (1991). Equality and Partiality (Oxford: Oxford University
Press).
Nagel, Thomas (2003). “Rawls and Liberalism,” in Freeman, 62–85.
Quong, Jonathan (2012). “Liberalism without Perfection: Replies to Gaus, Colborn,
Chan, Bocchiola,” Philosophy and Public Issues 2, 51–79.
Quong, Jonathan (2011). Liberalism without Perfectionism (Oxford: Oxford
University Press).
Quong, Jonathan (2013). “Public Reason,” in the Stanford Encyclopedia Online,
available at <http://plato.stanford.edu/entries/public-reason/>.
Rawls, John (1989). “The Domain of the Political and Overlapping Consensus,”
NYU Law Review 64, 233–55.
Rawls, John (1993). Political Liberalism (New York: Columbia University Press).
Raz, Joseph (1990). “Facing Diversity: The Case of Epistemic Abstinence,”
Philosophy and Public Affairs 19, 3–46.
Raz, Joseph (1998). “Disagreement in Politics,” American Journal of Jurisprudence
43, 25–52.
Rossi, Enzo (2014). “Legitimacy, Democracy and Public Justification: Rawls’
Political Liberalism vs Gaus’ Justificatory Liberalism,” Res Publica 20, 9–25.
Schroeder, Mark (2007). Slaves of the Passions (Oxford: Oxford University Press).
Schroeder, Mark (2008). “Having Reasons,” Philosophical Studies 139, 57–71.
Simmons, John A. (1999). “Justification and Legitimacy,” Ethics 109, 739–71.
Simmons, John A. (2010). “Ideal and Nonideal Theory,” Philosophy and Public
Affairs 38, 5–36.
142 David Enoch
Sobel, David (2009). “Subjectivism and Idealization,” Ethics 119, 336–52.

Sorensen, Kelly (2013). “Review of Nicholas Wolterstorff Understanding Liberal
Democracy: Essays in Political Philosophy,” Notre Dame Philosophical Reviews, available
at <http://ndpr.nd.edu/news/42380-understanding-liberal-democracyessays-in-
political-philosophy/>.
Statman, Dani (2009). “Critical Reflections on the Conscientious Exemption from
Military Service,” 31 Iyunei Mishpat (Tel Aviv Universtiy Law Review), 669–708
(in Hebrew).
Stemplowska, Zofia (2008). “What’s Ideal about Ideal Theory?,” Social Theory and
Practice 34, 319–40.
Vallier, Kevin and D’agostino, Fred (2013). “Public Justification,” in the Stanford Encyclo
pedia Online, available at <http://plato.stanford.edu/entries/justificationpublic/>.
Waldron, Jeremy (1987). “Theoretical Foundations of Liberalism,” The Philosophical
Quarterly 37, 127–50.
Waldron, Jeremy (1999). Law and Disagreement (Oxford: Oxford University Press).
Wall, Steven (2002). “Is Public Justification Self-Defeating,” American Philosophical
Quarterly 39(4), 385–94.
Wall, Steven (2013). “Critical Study: Public Reason and Moral Authoritarianism,”
The Philosophical Quarterly 63, 160–9.
Wolterstorff, Nicholas (2012). Understanding Liberal Democracy: Essays in Political
Philosophy, Terence Cuneo (ed.) (Oxford: Oxford University Press).
PA RT I I I
RIGHTS AND DUTIES
6
Territorial Rights
Justificatory Strategies
A. John Simmons
1. STATES AND TERRITORIES
States are defined in international law (for instance, in the 1933 Montevideo
Convention) as entities with permanent populations and fixed territories
under government control. I will focus here on just part of that definition: the
idea of states as territorial entities (indeed, as necessarily so, from this legal
viewpoint). Most of us do think of states in terms of their territories, first
learning about our own and other states by locating their colored territorial
shapes on maps. Exactly where the territorial lines defining these shapes are
drawn is obviously a matter of considerable importance to states, and the
history of the conflicts in which modern states have engaged has regularly
involved attempts, often successful, to redraw those lines by force.
Each modern state identifies itself with and claims a set of exclusive rights
over a particular territory.1 It is commonly assumed, I think, that these
claimed territorial rights are not merely legal rights; they are, at least in
decent or legitimate states, also moral rights over territories (or morally jus-
tified legal rights). It is on the possible moral justifications of states’ territo-
rial rights that I concentrate here. These rights include at least the right to
exclusive legal jurisdiction over the territory (that is, the right to coercively
regulate the conduct of all within the territory by means of enforcing all
legal rules and directives of the state); the right to full control over the land
and resources within (or constitutive of ) the territory that are not privately
1
I intend by this to deny neither the possibility of federal governance within states
nor the possibility of voluntary renunciation, transfer, or collective exercise of rights by
states. The rights discussed below should be thought of as those to which autonomous
nation-states take themselves to be entitled, prior to or independent of any voluntary
reduction or sharing of those rights within or between states.
146 A. John Simmons
owned; the right to tax and regulate privately owned land and resources
within the territory; the right to control or prohibit movement (of persons
and materials) across territorial boundaries; the right to determine the
standing of those within the territory (by, e.g., establishing rules governing
residency, diplomatic status, or citizenship); and the right to prohibit indi-
vidual or group territorial secession or alienation of territory to nonmember
persons or groups.2
Some of these claimed rights, notice, are primarily jurisdictional in
nature—that is, they are claims primarily to regulate and control a particular
territory, rather than claims to exclude persons from entering or using it.
Others of the territorial claims states make are more property-like, exclusion-
ary claims over a region, claims to choose who or what may pass over its
boundaries and who may use (and how they may use) the resources located
in it. I will not here discuss these more property-like claims, concentrating
instead on states’ jurisdictional claims over the territories with which they
identify themselves.3 The jurisdictional claims that states make are not, of
course, merely claims they make with respect to the geographical space
itself; they are primarily claims over the human beings located in that space.
States claim authority over those within their territories, the rights to make
and enforce laws and directives for them and the right to peaceful compliance
from them. States claim the right to “speak the law” to a set of persons—as
the literal meaning of “jurisdiction” suggests—and the relevant set of persons
is in the modern world primarily located in territorial terms. As a result, the
word “jurisdiction” now refers as well to the geographical area over which
legal authority is thought to extend.
I should emphasize that I am concerned here with the justification of
states’ claims over their particular territories, rather than with other kinds
of possible moral claims concerning group or state control over land and
resources (or about the modern state’s territorial form). For instance, it has
been regularly argued both that certain kinds of groups of persons have by
their natures a right to a territory on which to govern themselves—even
when there is no particular piece of land to which they obviously have a
valid claim—and that it is best in various ways for all or most of the land in
the world to be controlled by states or societies, rather than to be simply
open for all to use—even when there is no special reason why one state
rather than another should control any particular portions of the earth.
2
This is intended only as a list of the principal rights actually claimed by modern
states, not (as Cara Nine appears to suppose) a list of the necessary elements “of the con-
cept of territorial rights” (Nine [2012], 11).
3
I believe that the property-like rights modern states claim are in fact much harder to
justify than are their jurisdictional claims.
Territorial Rights: Justificatory Strategies 147
Groups’ rights of self-determination might indeed imply rights to territory;

and the need for societal stewardship, in the face of potential for “commons
tragedies,” might indeed show that the needs of all are best addressed by
some system of state control over land and natural resources. I set these
arguments (and other related ones) to the side. Similarly, I do not address
here arguments that states cannot perform their standard or core functions
without exclusive control over some territory or other.4 My concern is rather
with the possible justifications for particular states in the world to exercise
exclusive control over those particular geographical domains they claim as
their territories. And that issue cannot be adequately addressed simply by
showing that some system of state control over land is morally required, that
groups’ rights of self-determination imply their rights to control some land
or other, or that states cannot perform their core functions adequately with-
out control over some territory.5
My topic in this chapter, then, is the range of possible moral justifications
for states’ claims to exclusive control over particular geographical territories.
And my aim is to provide a structure for discussion of this topic by identi-
fying the various approaches, their salient strengths and weaknesses, and the
argumentative paths forward that present themselves from these analyses.
My hope, of course, is that the best path forward will become evident
from this discussion.
2. STRATEGIES
Philosophical theories of states’ territorial rights6 divide naturally into three

(very) broad types, each of which identifies a different sort of collective as
being entitled to the status of territorial rightholder.
Voluntarist theories maintain that groups of persons that choose to be
(and are capable of being) self-determining political societies in fact possess
the moral right to be or to make themselves autonomous states. Those
groups that make such a choice may have, in consequence, the right to be
4
E.g., “If we consider the range of functions that modern states perform, it quickly
becomes obvious that these functions cannot be carried out effectively unless the state has
authority over a determinate territory” (Miller [2007], 214).
5
In Chaim Gans’ nationalist theory of territorial rights, the distinction I make here is
identified as the distinction between “the right to territorial sovereignty” and “the loca-
tion of territorial sovereignty” (Gans [2003], 103).
6
Most of these theories are quite recent. When I first explored the topic of territorial
rights in print (in 2001), almost nothing had been written on the subject by contempo-
rary thinkers. The great historical political and legal philosophers also had very little to
say about such rights.
148 A. John Simmons
self-determining on the particular geographical territory that they occupy.

Exercising jurisdictional (and property-like) rights over that territory, the
argument goes, is necessary to their being genuinely self-determining. On
“plebiscitary” or majoritarian versions of voluntarism, it is the majority (or,
perhaps, a super-majority) of the persons living in some territory, acting
on behalf of all residents, that is taken to be entitled to make this choice.7
On Lockean-individualist versions of voluntarism (like the one I have myself
defended8), the territorial rights of voluntarily incorporated groups derive
not from the choices of the majority of residents in some territory, but only
from the choices of individual persons to convey to their states certain of
the rights they antecedently possessed over the specific land on which those
individuals live and labor.
Second, functionalist theories9 derive states’ rights to territorial control
from the fact that controlling territory is necessary to states’ performances
of their morally mandatory functions.
Because those functions must be performed territorially, a state’s right
to perform them implies a right to exclusive control over the particular
territories within which they are in fact performed. The moral mandates in
question are generally derived from either (broadly) Kantian or consequen-
tialist moral theories. Kantians take the morally mandatory function of
states to be that of “doing justice”—that is, making it possible for there to
be determinate, enforceable individual rights.10 Or they may (as in Rawls)
emphasize instead the need to guarantee that all basic goods in the society
are subject to a just distributive structure. Consequentialists take the mor-
ally mandatory task of the state to be that of maximizing overall good
results, such as happiness or well-being.11 In both cases, reasonably robust
rights over the specific territories in which the state does its work are thought
to be required for states to do their mandatory work.
Finally, nationalist theories12 hold that only groups that have certain
additional characteristics, beyond mere willingness to be a polity or effec-
tiveness in administering justice, possess the right of self-determination.
These characteristics are generally taken to include features like a shared
history, shared language, shared religion, or shared culture. And on many
7
E.g., Altman and Wellman (2009).
8
Simmons (2001).
9
This is Anna Stilz’ name for them in Stilz (2011), 576. See, e.g., Buchanan (2004),
370–1; Christiano (2006).
10
Stilz (2011), 581–2. Stilz’s basic position is Kantian; but her final account of states’
territorial rights, as we will see, is actually more of a hybrid.
11
The first apparent defense of a utilitarian theory of territorial rights is in Sidgwick
(1908), 252.
12
E.g., Gans (2003); Meisels (2009); Miller (2007).
versions of nationalism, a further characteristic that is required for a right

of self-determination is the group’s attachment to a particular geographical
territory on which the right to be self-determining may be exercised. The
territory in question might be the group’s “national homeland,” or it might
in some other way be specially tied to the group through the group’s his-
tory, its productive labor, or locations that have acquired national symbolic
value. But it is the specific relationship between a nation’s history and iden-
tity and a particular portion of the earth that is thought to ground a nation’s
territorial rights.
Thus, voluntarists argue that decent states’ territorial rights derive from
the moral importance of group or individual choice—controlling territory is
necessary to the success of any choice to create or to continue as a political
society. Functionalists say states’ rights to control certain territories are
required for them to perform the morally imperative tasks that only states
can perform. And nationalists justify states’ territorial rights through appeal
to the moral value of participation in cultural nationhood and to the ways
in which control over a particular territory is centrally implicated in what it
is to be one people, one nation.
While these three approaches to justifying states’ territorial rights disa-
gree with one another in reasonably fundamental ways, their defenders have
often cast their views as in even more basic opposition to a common oppo-
nent: the cosmopolitan.13 Cosmopolitan critics of the international state sys-
tem have argued that states’ boundaries ought to be “open” or very “soft”
(with respect to immigrants and travelers), being legitimately subject to
only quite limited control by states themselves. Similarly, cosmopolitans
have challenged states’ claims over the natural resources within their terri-
torial boundaries. Resource-rich countries, they argue, are morally required
at least to share their good fortune with less fortunate states, and possibly
to equalize access to or wealth deriving from their resources. Such argu-
ments, of course, are primarily attacks on the property-like territorial rights
claimed by states, leaving largely unchallenged states’ claims to their par-
ticular jurisdictional authority. Our question here will be precisely whether
theories of states’ justified territorial rights have the resources to defend
even the more “modest” jurisdictional claims made by decent states over
their particular territories.
Some recent theories of states’ territorial rights, of course, are not “pure”
versions of any of the three types just identified. It should, though, still
13
See, e.g., Pogge (1992), (1994). Left-libertarians are also generally skeptical about
the territorial claims made by most actual states; but I will not here discuss their position.
For a very brief summary of a left-libertarian view of territorial rights, see Steiner (2005),
34–6.
150 A. John Simmons
prove useful to identify the virtues and limits of the pure forms. For in
doing so we will be locating the argumentative “cores” of those attempts;
and we will then be better able to see the directions in which the failures of
the pure forms have pushed those attempts—and the directions in which
those attempts may need to be further pushed in order to successfully justify
strong territorial rights for actual states.
3. NATIONALISM
The most obvious strength of nationalist approaches to states’ territorial

rights is the ease with which they seem able to deal with the particularity of
such rights. Because many nations have historical, cultural, and emotional
ties to particular geographical territories, nationalists have a natural argu-
mentative avenue for explaining why particular territories—and not just
some territory or other—ought to be subject to the exclusive control of par-
ticular states. The “central case” used to motivate David Miller’s account, for
instance, is that of “a nation that over a long period occupies and transforms
a piece of territory and continues to hold that territory in the present.” That
nation, he claims, has a “quasi-Lockean basis” for a right to “the enhanced
value that the territory now has,” both in the “economic sense” and in terms
of “the symbolic significance” the national territory acquires.14 Gans relatedly
takes the central roles that certain territories can play “in the formation of
national identities” as an important determinant of the proper “location” for
peoples to exercise their rights of self-determination.15 The particular land
and the particular people shape and transform one another.
An equally obvious first difficulty for such approaches, of course, is that
even if we are persuaded of Miller’s conclusion in the idealized “central case,”
“other cases may lack one or more of [these] features . . . , so the strength of
the claim to territorial rights may vary.”16 Nations may not have occupied
their territory for very long, they may not have transformed it or enhanced
its value in interesting ways, or their identities may not have been shaped to
their geographical locales in the right ways. Worse, nations may not even be
the current occupants of the territories to which they in fact have these sorts
of connections. And most (if not all) of the states in the world that claim
robust rights over particular territories are not single “nations” in any very
strong sense, lacking the requisite cultural cohesion or national uniformity.
These facts immediately suggest that nationalist accounts of what justifies
14
Miller (2007), 217–18.
15
Gans (2003), 116.
16
Miller (2007), 219.
territorial rights simply may not apply to many actual decent states, threatening
to yield only skeptical conclusions.
Especially troubling are cases in which the state currently occupying (and
claiming) the territory is not the one—or the only one—with the appro-
priate sorts of historical and cultural connections to the land to trigger
nationalist-style reasoning about territorial rights. And, of course, these
cases are most troubling when the current occupant took possession of the
territory in question by plainly unjust or illegitimate means. Such cases are
unhappily commonplace, so any nationalist account that hopes to apply its
arguments (non-skeptically) to the real world is obliged to address them.
The standard argumentative move is to simply claim that the rights of inno-
cent peoples (and persons) who are illegitimately annexed, conquered, or
expelled “fade out” with the passage of time, while new rights for those who
have wrongly seized their territories (or for their descendants) “fade in.”
While it is, of course, hard to be very precise about this process, it is a pro-
cess that is assumed by many (including non-nationalist) writers on the
subject to reflect the moral facts. Miller, for instance, denies that his posi-
tion amounts to “a charter for thieves,” but allows that the question of who
eventually “has the better title will be a matter of judgement.”17 So while the
wrongs states and groups do in achieving their territorial goals may initially
call for restitution, for a restoration of the status quo ante, once those wrongs
become older, the wrongs are “superseded”18 and there springs into being a
new set of moral rights for a new set of people to control the unjustly
acquired territory.
Because virtually no modern states can make territorial claims that are not
historically stained by such injustices, it may seem that a view like Miller’s
is a necessary feature of any adequate theory of territorial rights. We should
note two points, however. First, the devil here is surely in the details. Exactly
when and why rights go away and appear, how soon and for what reason
victims lose their claims to restitution or reparation, and wrongdoers (or
their heirs) acquire claims to ill-gotten gains, is a matter of significant theo-
retical and practical importance and great theoretical difficulty. Second, we
will only be required to accept a very dramatic and suspiciously self-serving
account of the supersession of our states’ territorial wrongs if we think a
standard of adequacy for theories of territorial rights is that they not be
interestingly revisionist in their implications—that is, that they be able to
explain why stable, reasonably just states in the world actually have legiti-
mate territorial jurisdiction over all of the territory that they claim as theirs.
If we are prepared to accept instead the possibility that even reasonably just
17
Miller (2007), 220, 219.
18
Waldron (1992).
152 A. John Simmons
states may not be morally entitled to all they claim, we can defend a more
plausible account of the moral significance of the many relatively recent
wrongs done by states in their quests for territory. For now, however, I want
simply to flag this problem confronting nationalist accounts—call it “the
problem of rights supersession”—since it is, I think, both a serious problem
and a problem that (as we will see) also confronts many non-nationalist
accounts of states’ territorial rights.
Surely, though, the most severe hurdle faced by nationalist accounts of
territorial rights is simply the absence of cultural/national uniformity within
the marked boundaries of the political world. Nationalists tend to locate the
relevant territorial rights in the cultural or national majorities within plural-
istic political units, leaving the preferences, interests, and goals of those not
included in the majority national group disturbingly to one side. Miller’s
response to the problem is to claim that while all residents of a national
territory may not share all of the same goals, they all do have “a common
interest in being able to set those goals through democratic debate” and in
belonging “to a community with a shared sense of national identity.”19 But
this response seems only to push the relevant difficulty one step back.
Minority groups and individuals may well have a version of the interests
described by Miller. But they might well have no interest at all in a shared
setting of goals, engaging in democratic debate, and achieving a shared
national identity in the particular political setting in which they find them-
selves—a setting in which they will be systematically outvoted on matters of
central concern to them by the majority national group. Call this difficulty
“the problem of trapped minorities.” It is again, as we will see, a difficulty
shared by a variety of non-nationalist approaches.
4. FUNCTIONALISM
The strongest point of nationalist approaches—namely, their ability to

explain and justify the particularity of territorial rights—seems a significant
weakness of functionalist approaches. Legitimate states for the functional-
ist, remember, are simply functional political units: that is, institutionally
structured collections of persons, of whatever size or location, that success-
fully perform their morally mandated functions (such as administering justice,
establishing social equality, or adequately promoting social utility). It is thus
the institutional structure of the state—and the way that structure operates
in the lives of its citizens—that matters from the functionalist viewpoint,
not the location or the history of the state. The fact that functional states
Miller (2007), 224.

19
happen to arise in one place or another need not reflect any special relation-
ship between those states and the land they occupy. If those states could
function effectively elsewhere or function effectively with altered bounda-
ries, simple considerations of their morally mandated functions present
no principled bar to such changes. Without additional, non-functionalist
principles in play, it is hard to see why our relocating, adding to, of sub-
tracting from a legitimate state’s territory would constitute a wrong. Further,
of course, currently functional states can plainly rest on a sordid history of
wrongdoing. What matters for the functionalist is that the state in ques-
tion here and now successfully administers justice or successfully promotes
social happiness.
This means, of course, that functionalist theories will not only have prob-
lems with the particularity of territorial claims, they will also face the same
problems of trapped minorities and rights supersession that face nationalist
theories. States can perform their morally required functions even with
unwilling parties and groups subjected to political authority within their
borders; and the functionalist, like the nationalist, owes us a convincing
account of when and why the rights of wronged parties just fade away as
they are opposed by the interests of established, adequately functioning
states. These difficulties for functionalism add up to what I have elsewhere
called the “the boundary problem.”20 Pure functionalism simply cannot
guarantee that the boundaries of the territories counted by the theory as
rightfully subject to state authority do not enclose people who have plainly
been subjected illegitimately to states’ coercive power. Because functional-
ism grounds states’ territorial rights in their current provision to areas of the
services that states are morally required to provide, states can acquire justi-
fied territorial rights, according to the logic of the functionalist approach,
simply by making themselves the provider to a territory of those services.
Suppose the United States one night secretly moved its southern border
barriers a few miles further south into Mexico, claiming the newly enclosed
Mexican territory and the surprised Mexican subjects living on it as its
own.21 There would appear to be nothing wrong with this according to
functionalist reasoning—provided only that the United States extended its
effective administration of justice to this new territory as well. It is a state’s
effective administration of justice over a territory and people, not the history
of the state’s acquisition of territories and subjects, that functionalism iden-
tifies as the source of its legitimate territorial rights.
Simmons (2013).
20
I am supposing further, for purposes of this example, that both states are sufficiently
21
just (or possess whatever other legitimating properties are necessary) for the relevant
functionalist view to count both as having started with legitimate territorial rights over
their claimed territories.
154 A. John Simmons
This is a problem that is not faced (at least as immediately or directly)

by the alternatives to functionalist theories of territorial rights. Nationalism,
for instance, grounds the state’s territorial rights in group “attachments” to
particular territories. While national groups (and their territorial attach-
ments) may, of course, extend across existing political boundaries, national
groups at least cannot simply “make themselves” attached to some terri-
tory outside their current domain.22 Similarly, political societies cannot,
on the Lockean model, simply claim land beyond that which is lived and
labored on by those societies’ willing members. Functionalism, however,
seems unable to avoid licensing plainly illegitimate acquisitions of territo-
rial jurisdiction.
I consider at length in section 8 the question of whether a largely func-
tionalist theory can plausibly address these problems through the addition
of various non-functionalist principles. And I will consider more fully
there as well both the general motivations for functionalist approaches to
the problem of states’ territorial rights and what I take to be the limits on
hybrid, pluralistic approaches to this subject.
5. VOLUNTARISM (MAJORITARIAN)
The majoritarian, plebiscitary version of voluntarism, though apparently

designed to avoid them, can in fact be seen to share many of the same
problems faced by the nationalist and functionalist approaches to territorial
rights. According to Altman and Wellman’s version of the view, a legiti-
mate state is a “territorially based,” “nonconsensual form of association” that
enjoys a group “right of self-determination.”23 Any group that is “sufficiently
large, wealthy, politically organized, and territorially contiguous so that it
can secure for all individuals in the territory the essential benefits of political
association” (while respecting the rights of others), has the right to form or
sustain its own state on, and exercise robust control over, the territory it
occupies.24
Why even describe as “voluntarist” an account on which legitimate states
are characterized as “nonconsensual associations”? This account makes legit-
imacy and territorial jurisdiction a matter of the choice of the relevant (polit-
ically capable) group to exercise the core functions of the state. But the
22
There is, however, no very obvious reason why nations could not form bizarre
attachments to (or otherwise build their identities around) distant territories or land-
marks with which they had little or no actual history of physical interaction.
23
Altman and Wellman (2009), 4–5.
24
Altman and Wellman (2009), 46–7.
“choice” at issue, of course, is the choice of the group conceived as a territo-

rially organized whole. And that position naturally suggests the question of
whether it can plausibly solve “the problem of trapped minorities.”
It may seem that the answer is obviously “yes”, since “trapped minorities”
can (according to Altman and Wellman’s theory) escape their traps by opt-
ing for secession, choosing to create legitimate states of their own on their
own territory. While this will be no help to “trapped” individuals or to
trapped small, dispersed, or disorganized groups,25 any substantial, non-
impoverished group appears to have a reasonable route out from the trap of
permanent minority status. Any group that is “willing and able” to perform
as a legitimate state may do so (at least provided existing legitimate states are
not disabled or unfairly impoverished by this choice). But consider for a
moment what “willing and able” actually means here. “Willing,” of course,
refers to the will of the majority of residents. Suppose, however, that you are
untroubled by how this simply pushes one level “downward” the problem
of trapped minorities (for example, the problem of Union sympathizers
trapped in the seceding Confederate States of America). Focus instead on
what it means for a group to be “able” to function as a legitimate state. As
Altman and Wellman understand it, this means, first, that the group in
question must be territorially concentrated. Second, it means that the
group must be “sufficiently large, wealthy, and politically organized” to act
as a state.
But notice that groups can be made or kept small, poor, politically disor-
ganized, and territorially dispersed by the (wrongful?) coercive actions of
other parties (such as other states or groups that oppose their political inde-
pendence). The will to act as a legitimate state amounts to nothing if it is
defeated by force at every point. So the “trap” in question will certainly look
more dire and unavoidable if others can legitimately simply use coercion to
guarantee that it will not be escaped. Indeed, even groups that are territori-
ally concentrated, large, rich, and organized can still be stopped from acting
as legitimate states if other states or groups are “willing and able” to forcibly
intervene to stop them from doing so. Does a group fail to count as rele-
vantly “able”—and so fail to have a right to self-determination as an auton-
omous state—if some other state will simply use force to prevent any
attempt by the group to function as a state?
Altman and Wellman, I assume, want to answer “no” to this question. So
suppose that we count as “able” to function as a legitimate state all those
groups that could do so if others did not coercively intervene to prevent
their doing so? Then, however, we must ask several questions: first, how far
25
All are free to exit, of course, but only by abandoning their lands and subjecting
themselves to the dominion of whatever other society (if any) is willing to take them in.
156 A. John Simmons
back in history is coercive intervention by others going to “count” in deter-

mining this? Consider a group that could have satisfied the requirements for
being “willing and able” to act as a state, but that was forcibly expelled and
dispersed by a militarily superior power—such as the Acadians expelled
from Nova Scotia by Great Britain (during its conquest and resettlement of
French Canada). Once they were dispersed, the Acadians were no longer a
territorially concentrated or politically organized group, and their original
territory was rapidly settled by others. Did the Acadians still possess the
right to be a state? If so, on what territory did they have a right to establish
their state? If not, then Altman and Wellman’s theory simply privileges the
existing territorial concentrations of persons, for no good moral reason and
regardless of how those concentrations were achieved.
But if such wrongs must be righted—and if the Acadians still possessed the
right of self-determination at issue, even though “unable” to exercise it—when
(if at all) did those wrongs and rights “fade away”? This, of course, is just “the
problem of rights supersession” again, and Altman and Wellman must solve it
before their position can be satisfactorily defended. The success or failure of a
group to acquire the characteristics that give it the right to be a state and to
control the territory it occupies (according to plebiscitary voluntarism) in each
case has a history, and that history may be filled, even quite recently, with pal-
pable wrongs. Unless the theory can convincingly address that fact, it will
continue to appear simply without argument to privilege the status quo.
Indeed, the plebiscitary voluntarist theory builds into its basic logic an
obvious privileging of the territorial status quo. If a territorially concen-
trated group derives its right to be self-determining on that territory from
majoritarian voting within the group, the theory simply must be assuming
that existing territorial concentrations of persons constitute pre-legitimated
groups.26 Otherwise, there is no reason to suppose that the will of the
majority has any right to determine the lives or constrain the choices avail-
able to all persons within that group. Majority voting within the group of
six-foot-tall-men, say—a group in which I happen to be included—surely
(and happily) doesn’t establish the right of the majority of six-footers in the
world to impose their preferred arrangements on me. Why, then, should we
suppose, with the plebiscitary voluntarist, that the majority of persons who
just happen to live in some arbitrarily defined geographical territory have
the right to create a self-determining political society with territorial jurisdic-
tion and political authority over the minority of residents? We should sup-
pose this only if we believe that persons come pre-sorted into pre-legitimated
groups, groups identified only by their members’ geographical proximity to
one another and bounded only by whatever arbitrary outer geographical
26
See Nine (2012), 58.
boundaries we might choose to identify. But relying on that belief at the

theory’s foundations plainly just begs all of the important questions about
what justifies states’ claims to territory.
Wherever mere majority will is allowed to dominate the determination
of state territorial boundaries, the manner of composition of the body of
which that majority is the majority should be our primary moral concern. Just
as my (populous) neighborhood may not legitimately incorporate the adjoin-
ing (less populous, less affluent, less organized) neighborhood without its
consent and then control it by majority rule, political bodies may not legiti-
mately subject to their authority all the unwilling people that they are able to
surround and outvote. Nor should it matter that those so subjected were
incapable of themselves functioning as a political society, especially if this
incapacity is simply accepted as such without consideration of its causes. If a
group’s “inability” to function as a state is understood independently of the
history and source of that inability, plebiscitary voluntarism will simply face
the same kind of “boundary problem” that is faced by pure functionalism.
Altman and Wellman’s only apparent attempt to counter such concerns
occurs in their response to (what they call) the “regress argument.” That
argument maintains, against their view, that one can’t use the principle of
self-determination “to determine political boundaries, unless one first decides
what the boundaries are within which voting is to take place. But the deter-
mination of the boundaries within which voting is to take place is itself a
determination of political boundaries.”27 They consider this problem specif-
ically in connection with the issue of secessionist movements; but I think
the problem is one that confronts their view throughout.
Unless we antecedently assume that existing occupations of territories by
groups are legitimate and uncontroversial, how can we possibly proceed to
argue that voting within those current “group boundaries”—that is, within
the bounded territories simply claimed by or occupied by groups—will
ground genuine moral rights to self-determination?
Altman and Wellman’s response to the argument is that it is “possible to
stop the regress in a nonarbitrary way . . . ‘Let the separatist movement spec-
ify the area within which the plebiscite is to be held’.” In other words: how
can groups desiring political autonomy complain if they are themselves per-
mitted to identify the boundaries of the territory within which a vote on
independence will be conducted? Altman and Wellman allow that “of
course the precise contours of the territory picked out by the separatists is
arbitrary in some respects.”28 But they appear to regard that problem as,
relatively speaking, just a minor matter of detail.
27
Altman and Wellman (2009), 49.
28
Altman and Wellman (2009).
158 A. John Simmons
This seems to me, however, a far deeper and more serious problem than
their response acknowledges. Both the boundaries of the original states and
the boundaries picked out by those forming new states may be morally
arbitrary (or deeply illegitimate) in very important ways. For instance, “the
separatists” discussed by Altman and Wellman might easily include in their
“specified” territory not just the land occupied by their supporters, but also
land occupied by others (who will be easily outvoted by those supporters),
possibly because that land contains valuable resources or other desirable
geographical features—just as the territory from which they desire to sepa-
rate may itself have been formed by “trapping” unwilling minorities in var-
ious ways. Provided only that this “specification” does not incapacitate their
original state (or take an unjust share of the state’s resources, say), there
appears to be no bar in the theory to such majority choices by secessionists.
While Altman and Wellman do condemn the annexation of less popu-
lous by more populous states,29 they do so only where the less populous
groups are already organized as states. Those people and groups who are
“unable” to function as states (for whatever reason?) are simply left as grist
for the statist mill.30 Altman and Wellman seem primarily concerned to
affirm that unwilling individuals, sprinkled here and there throughout an
otherwise willing and territorially contiguous group, may be legitimately
subjected to political authority without their consents. But the wrongs their
position permits—in both terms of trapped minorities of significant sizes
and rights superseded in an instant in the interest of the present possessor of
territory—seem vastly more significant than those they seem principally
concerned to deny. And these difficulties push the theory in a more fully
voluntarist direction, toward a view capable of greater sensitivity to histori-
cal wrongs against peoples and persons.
6. VOLUNTARISM (INDIVIDUALIST)
So the standard contemporary approaches to justifying states’ territorial

rights all seem to share a set of apparent problems: their justifications of
territorial jurisdiction for reasonably just states permit those states to trap
unwilling individuals and minorities within those jurisdictions, and they
appear to deny persons and groups that are plainly wronged in the process
29
Altman and Wellman (2009), 52.
30
Wellman does elsewhere insist that “willing and able” political groups “who occupy
a territory enjoy jurisdictional rights over this land” only “other things being equal.” But
the only example he gives of a case where other things are not equal is again that of the
forcible annexation of one state by another (Wellman and Cole [2011], 102–3).
of territorial acquisition and concentration any obvious just remedy for

those wrongs. This appears to be true whether the theory attempts to derive
territorial rights from the value of collective free choice and self-determina-
tion—as on the plebiscitary voluntarist view—or from the value of national
orientation of a people on their land—as on the nationalist view—or from
the value of justice or social happiness—as on the functionalist view.
Consider now a much older view of states’ justified territorial rights that
is distinctive precisely in virtue of its effectively avoiding all of these prob-
lems that are shared by its competitors. The Lockean view—the individual-
ist version of voluntarism—rests on three compelling claims. First, that the
only persons who are legitimately subject to a state’s authority are those who
are willingly subject to it. Second, that the only clear cases of rightful state
claims to territorial rights (over particular geographical areas) are claims to
the territory lived on and labored on by that state’s willing subjects. And
third, that the rights of those persons who are wronged in states’ territorial
acquisitions do not simply fade away in the interest of the more powerful or
the more numerous.
Locke himself, of course, argued that only those who have consented to
membership in a political society are legitimately subject to its authority
and have the obligations of citizens. And those who agree to be citizens,
Locke claimed, must be understood to consent as well to submitting to the
state’s jurisdiction any land over which they have rights of occupancy or
ownership. The heart of each legitimate state’s rightful territory is thus con-
structed piecemeal from the free choices of persons to submit both them-
selves and the land on which they live and work to the state’s authority. The
Lockean view condemns both the political subjection of the unwilling and
the exercise of territorial control by states over areas not legitimately
employed in their willing subjects’ lives.
The principal virtues of Lockean voluntarism are plainly these: first, it
explains the particularity of states’ territorial claims in a natural and straight-
forward way, by identifying each state’s territories with the particular areas
in which its willing members live and labor. Second, the Lockean theory has
simple and persuasive answers to the problem of trapped minorities and to
concerns about the supersession of rights and past injustices done by states.
The Lockean view is committed to the position that states may do no more
to restrict freedom of the unwilling located within their boundaries than
private persons would be entitled to do to one another in a state of nature.
Trapped minorities may not be subjected to political authority without
their consent—though they may, of course, be watched and defended
against—because they enjoy the same rights to live and choose freely that
each polity’s members exercised in choosing to create or join a state. Just as
such minorities may not undermine the rightful political arrangements of
160 A. John Simmons
those around them, those whose arrangements they are may not interfere
with the rightful activities of the unwilling.
Similarly, Lockean voluntarism rejects the idea of simple supersession of
rights by seizure or mere passage of time (moral rights being, on this view,
“imprescriptible”).31 It cannot be embarrassed by theoretical insensitivity
to the plights of the expelled, the annexed, and the wrongly subjected—
even when those unfortunates are mere individuals or when, as groups, they
have never been able or been permitted to organize politically. The Lockean
view is thus not vulnerable to charges of over-eagerness to simply affirm
the legitimacy of the territorial status quo. The rights of those maimed in
the machinery of politics do not simply fade away for the convenience
of the powerful or the numerous. Those whose rights have been violated in
creating or reshaping states retain the right to rectification of those wrongs.
Rights supersession is, according to the Lockean voluntarist position, nor-
mally just wishful thinking, typically done by those who most stand to
benefit from it.32
With all of these noteworthy virtues, why have political philosophers and
political theorists not jumped at the opportunity to embrace the Lockean vol-
untarist theory? The answer is not, I think, that proponents of non-Lockean
views simply reject the powerful intuitions on which the Lockean theory is
built. For it is at precisely the point where concerns about historical wrongs are
brought clearly into focus that most rival theories are motivated to take hybrid
form, introducing historical principles in order to avoid the strongly counter-
intuitive implications of their core theories. I take, instead, the most persistent
and fundamental objections to Lockean voluntarism to be three.33 The least
31
Lockean voluntarism must still develop a theory of “moral residue” for cases where
rightholders and wrongdoers have died with neither rectification of the wrongs nor
inherited rights and duties of rectification. I know of no satisfactory theory of that sort.
But all theories of territorial rights are equally in need of one.
32
Most Lockeans (including myself ) hold, as well, that rights over land are heritable
and do not simply disappear as the specific rightholders and wrongdoers die off. And
rights over land held by freely incorporated groups continue to be held by those groups,
even as their memberships change.
33
I leave to one side an equally common basis for objection—namely, the defects in
the arguments advanced by Locke himself. Stilz, for instance, appears to (mistakenly)
reject the Lockean approach to territorial rights almost entirely because she (correctly)
takes Locke’s own arguments to be defective. Locke’s conservative ambitions—to justify
continuous, enduring rights over their full claimed territories for non-tyrannical states—led
him to compromise his own foundational principles and to make some of the errors Stilz
identifies (Stilz [2009], 192–4). But a more genuinely voluntaristic (and less conservative,
more revisionist) Lockean position than Locke’s own, one that takes seriously each per-
son’s equal right to a fair share of the earth and its resources, is surely a better test of the
Lockean approach to these issues—just as Kantians (like Stilz) often find their own posi-
tions more defensible when they depart from the letter of Kant’s texts (Stilz [2009], 198
[note 8], 203).
troubling of these, despite its surprising resilience, is the claim that the Lockean
account confuses the idea of property with the quite different idea of juris-
diction or territory, in consequence of which the Lockean account of states’
territorial rights must be equally confused.34 To identify the two would, of
course, be confused. But property ownership clearly has a jurisdictional
aspect, just as territorial rights have a property-like aspect (making it not at
all confused to suppose that the latter might derive from the former).
Landowners have (“jurisdictional”) rights to make certain kinds of rules
to govern their lands, thus unilaterally restricting the liberty of those who
choose to enter on their land. Landowners who choose to submit their land
to a state’s authority give the state’s rules priority on their lands, thus
accounting for the state’s consequent jurisdictional rights over those lands.
They also agree to allow the state to regulate their land in other ways (which
they were originally entitled to do themselves), including controlling those
boundaries of it that will constitute parts of the state’s territorial boundaries.
The result is a kind of sharing between state and subject of the various inci-
dents that comprise full rights in land, and a concentration of some of those
incidents (received from all subjects) in the hands of the state. This latter
concentration, I think, is an accurate representation of our normal concep-
tion of states’ rights over their territories, not any kind of confusion of ter-
ritory with property.
Often this first objection is combined with a second, in whose light the
first becomes more intelligible. If rights in land are necessarily created
through legal or institutional rule, then it plainly makes no sense to claim
that individual rights in land serve as the justifying foundation for these
institutions (i.e., the institutions that are responsible for creating rights in
land in the first place). The first objection thus really appears often to rest
on the second: that it is not possible to make sense of rights in land outside
of a state whose laws establish them; so we naturally can make no sense
either of grounding a state’s territorial rights in the subjection (by willing
members) of private rights in land to that state’s jurisdiction. Notice, how-
ever, that the required skepticism about pre-legal rights in land must extend
further than mere skepticism about Lockean “natural” property rights. It
must extend as well to all conventionalist theories of rights in land (such as
Hume’s), according to which extra-legal moral rights in land and chattels are
possible given the establishment of appropriate interpersonal conventions
to define and regulate them. If extra-legal property in land is possible—be
it naturally or conventionally grounded—then such land can be subjected
to a state’s jurisdiction, as the Lockean model requires for justified territorial
rights. So this objection must claim quite a lot about the impossibility of
See, e.g., Nine (2012), 73–6; Buchanan (2003), 232–4.

34
162 A. John Simmons
private rights in land outside the state, requiring somewhat more philo-
sophical nerve than at first might seem to be the case.
The third and undoubtedly most fundamental obstacle to the acceptance
of Lockean voluntarism, however, is that the theory offers an account of
states legitimate territorial rights that does not match up very well with the
real-world claims made by actual modern states.
States are not voluntary associations, nor have the territories states’ claim
been constructed from the submission of land by willing subjects to state
jurisdiction. Lockean theory thus seems to describe an extremely demand-
ing ideal that is simply too distant from the hard reality of the world to
permit its application to real-world territorial claims and disputes.35 The
plebiscitary voluntarist, functionalist, and nationalist accounts may appear
to do better on this score, since they are focused more on states’ present
characters and capacities and less on the historical processes that produced
the present distribution of states’ claimed territories. These views thus seem
more immediately applicable to real-world territorial claims and disputes,
despite their shared theoretical defects (described above), and less likely to
require dramatic revisions in our views about the actual territorial rights of
real-world states.
In part, this complaint about the Lockean theory is simply correct.
Because it takes very seriously historical wrongs and the rights of the unwill-
ingly subjected, the theory’s practical implications will inevitably be more
revisionist than will be those of the alternative theories I’ve criticized here.
But all of these theories are ideal theories; all describe ideals to which real
states’ territorial claims will conform only very imperfectly. Many real states’
territorial claims will not qualify as legitimate on the terms of plebiscitary
voluntarism, functionalism, or nationalism, nor will these theories obvi-
ously yield clear solutions to the most pressing of the world’s actual territo-
rial disputes. These are, after all, philosophical theories.
The job of philosophical theories of this sort, in my view, is to identify
our moral target, to describe how states’ territorial claims could be fully
rightful and legitimate. There will then be separate and difficult questions
about how, in a distinctly non-ideal world, we can approach that moral
target in an efficient and fair fashion. That will involve, first, determining to
what extent territorial claims made by real-world states count as legitimate
according to the ideal theory; and, second, using the ideal to identify the
most serious wrongs done by states (and their subjects) in establishing their
acknowledged territorial boundaries. Only then can we meaningfully proceed
35
Miller, for instance, rejects such views as “impossibly demanding . . . , putting virtu-
ally all borders into question” (Miller [2007], 220).
with the business of recommending practical policies that best correct these
wrongs, beginning with the most serious, and always focused as well on
practical and moral limitations on required reforms.
The true test for an ideal philosophical theory, in my view, is not how
closely and comfortably its prescriptions match the ways in which we actu-
ally live our political lives, but rather how plausibly it identifies the most
grievous kinds of wrongs that we do each other in the course of those lives.
And Lockean voluntarism, I submit, identifies the wrongs that need right-
ing in a clear and compelling way, putting us in position to attempt to
redress them and to gradually achieve a more rightful condition. For the
Lockean, our political ideal must be a world in which each person is, as fully
as possible, treated as a self-governing equal.
The alternative theories I’ve criticized are theoretically insensitive to too
many of the clear wrongs that states do in our names. When states insist
on exercising jurisdictional authority over land occupied by unwilling
subjects, or when they expel or destroy the innocent in their quests for
territorial control, they wrong persons in ways that require rectification,
either through the adjustment of territorial boundaries or through genu-
ine reparations of some other sort. Indeed, I think that modern states also
frequently do wrong simply in their efforts to control the sites of a wide
range of the world’s natural resources (through their more property-like
territorial claims).
In real-world disputes about territorial boundaries or about rights to nat-
ural resources, facts will inevitably be contested, claims will inevitably con-
flict, and disagreement will inevitably persist. A sound philosophical theory
of territorial rights can at best only identify the salient moral vectors in play
in such disputes and lay out the general guidelines for pursuing the best
resolutions of them.
7. ALTERNATIVES?
Do other alternatives remain? Here I will merely mention two possible

strategies that might appear to circumvent the problems I’ve identified
without yielding to the revisionist Lockean approach. The first employs an
approach of “indirection.” Instead of trying to directly justify particular
distributions of territories to states (by appealing to some nationalist, func-
tionalist, or voluntarist theory), we might try to justify these distributions
indirectly—by justifying the international legal order that has made or con-
firmed (much of ) the extant distribution (through its acceptance and its
defense of [many] existing territorial boundaries).
164 A. John Simmons
The problems with such an approach, however, should be obvious. If

the international order is not itself well justified, it cannot convey through
its history any legitimacy to its recognition of territorial claims by states.
And the international order has been and continues to be indefensibly
structured around the interests of powerful states, while its lack of enforce-
ment mechanisms undermines its capacity to uphold any judgments of
illegitimacy it makes (judgments that are typically forgotten after a suita-
ble period of mourning). The international order has permitted states,
individually or collectively, to lay claim to all of the useable universe,
including that which is unknown and unused (such as outer space and the
inner earth). A “legal order” with such a history lacks, in my view, suffi-
cient legitimacy to convey legitimacy to the territorial claims by states
that it upholds.
Second, it might seem that the problems faced by existing theories of
territorial rights could be neutralized if those theories were understood not
as ideal theories, but as non-ideal theories of states’ territorial rights. A non-
ideal theory could acknowledge, for instance, that the subjection of trapped
minorities or the non-rectification of past wrongs constitute genuine injus-
tices, while still arguing that accepting (at least much of ) the existing distri-
bution of particular territories to particular states is the best course of action
under our current morally non-ideal conditions. Now, exactly how we should
understand the nature of non-ideal theory in moral and political philos-
ophy is, of course, far from settled. But my own view is that we should
understand non-ideal theory as a transitional theory of justice (or right-
ness). The rules or principles of non-ideal theory, so understood, would
be rules designed to direct us on the best (i.e., the most efficient, feasible,
but still morally permissible) route to achieving full compliance with the
right.36 And if we understand non-ideal theory in that way, I do not think
it can be plausibly maintained that simply continuing on our present
course is at all likely to bring about full compliance with ideal prescriptions
governing states’ territorial claims. I have just referred above to my beliefs
about the deficiencies of the international legal order in the territorial
claims that it upholds. This order has always plainly been, I think, far more
clearly oriented toward achieving a stable compromise between the inter-
ests of the great powers than it has been toward achieving full compliance
with any ideal conception of the right. If so, non-ideal theory is unlikely to
dictate anything like simple compliance with that order—and with its dic-
tates with respect to states’ territorial rights—as the best transitional route
to perfect justice.
36
I defend this view of the relationship of non-ideal to ideal theory in Simmons
(2010a).
8. PLURALISM
As I argued above, the standard reasons for hasty rejection of the Lockean
voluntarist position—on both territorial rights and political authority—are
less compelling than they might initially seem. Can nationalist, functional-
ist, or plebiscitary voluntarist theories effectively avoid these perceived
“costs” of embracing Lockean voluntarism by dealing in other ways with the
problems the Lockean view allows us to avoid, such as the problems of
trapped minorities, and of awkward commitments regarding the superses-
sion of rights? Perhaps those theories could simply be converted into more
pluralistic, hybrid accounts by adding a set of new principles designed to
correct their core theories’ inabilities to effectively address these problems. I
will here briefly examine just one recent effort to “hybridize” an alternative
theory of territorial rights: namely, Anna Stilz’s Kantian theory of territorial
rights. The problems with Stilz’s account are, I think, representative of those
facing hybrid accounts generally.
According to Stilz, a state enjoys “a prima facie claim to territorial juris-
diction” over a particular geographical territory if, first, “the state is in fact
legitimate” (i.e., it protects “freedom-as-independence to a sufficient degree”)
and, second, the state’s “legal system defines property rights over [that] par-
ticular area of land.”37 The “extent” of states’ jurisdictions and the particular
“locations” of their territorial rights thus at first appear to be determined
simply by the practical “reach” of legitimate states’ institutions for adminis-
tering justice. Legitimate states are morally entitled to govern whatever par-
ticular territories they actually effectively govern. So far, this is just the kind
of account that we see in other Kantian accounts of territorial rights. After
all, what appears to matter centrally in the Kantian project in political phi-
losophy is that freedom be secured by the subjection of all to effective jus-
tice-administering institutions, not that individuals be subject to this or
that particular political/legal administrative structure. So the problem of
identifying the particular territories over which states have de jure territorial
rights is allowed to be settled according to the location of the de facto terri-
torial control exercised by legitimate polities.38
Stilz, however, correctly recognizes that “if this were all that mattered . . . there
would be no objection to annexation,” since the annexing state might be
able to administer justice as well as (or better than) the existing domestic
Stilz (2011), 587, 590.

37
In light of its distinguished proponents, this probably qualifies as the “standard”

38
view in contemporary political philosophy: the boundaries of existing states are “acciden-
tal” or “arbitrary,” but they are not so in any way that challenges their legitimacy. See, e.g.,
Rawls (1999), 39; Dworkin (2011), 381–2.
166 A. John Simmons
institutions of the annexed state.39 Even structurally legitimate states can

wrong persons or groups by “over-reaching,” by extending their institu-
tional control in wrongful ways.40 Thus, Stilz allows that a legitimate state’s
prima facie territorial rights can be “defeated” if the particular territory in
question (that is, some particular portion of the territory which the state
effectively governs) has been acquired by wrongfully displacing other groups
from that territory or by wrongfully annexing other groups along with their
territories.41
The wrongs involved in the two “defeater” cases are, according to Stilz,
different but related sorts of wrongs. The wrongs involved in cases of wrongful
displacement are violations of individuals’ “rights of occupancy.” Individuals
have not only the right to be somewhere, but the right to be in that particular
territory in which legal residence “is fundamental to the integrity of [their]
structure of personal relationships, goals, and pursuits.” Each individual has
a right that others respect “the centrality of territorial occupancy for an
individual’s personal autonomy—his ability to form and pursue a concep-
tion of the good.”42 By contrast, the wrongs involved in cases of wrongful
annexation are violations of collective rights of the peoples in the annexed
territories—specifically, their rights to an analogous but collective autonomy.
Even when peoples’ states disappear, the peoples may persist, along with their
right to the particular legitimate political institutions “they have created
together through their political history.” And even when annexation does not
involve destroying an existing government, it may still violate a “residual
claim” for a people “to reconstitute legitimate political institutions on their
territory when their prior state fails, becomes illegitimate, or is usurped.”43
Both of these accounts (of wrongs that states may do in attempting to
extend their territorial jurisdiction) seem to me troubling. We can begin
with wrongs of displacement. Because Stilz analyzes the wrong involved
here as a violation of individuals’ rights of occupancy—and because she
defends those rights as based in each individual’s “need for a stable legal
39
Stilz (2011), 595.
40
See Simmons (2010b), (2013).
41
Stilz (2011), 590.
42
Stilz (2011), 583, 585. It thus remains unclear from Stilz’s discussion how, if at all,
we are to explain the (apparent) wrong involved in coercively relocating faultless people
whose goals, projects, and so on are not specially connected to the place in which they
legally reside. In her more recent work on occupancy rights, it appears that the only
wrong at issue in such cases may be the generic wrong of coercing people without justifi-
cation (Stilz [2013], 340–1). But if relocation does not involve violating an occupancy
right, justifying relocation may turn out to be quite easy. On the Lockean view, by con-
trast, persons have enduring rights over all of the land (within their fair share) of which
they make productive use.
43
Stilz (2011), 591, 595.
residence”—Stilz’s account requires “a certain balancing” between the com-

peting occupancy rights of displaced persons and those of persons who sub-
sequently (and “without fault”) build their lives on the wrongly seized
lands.44 But the actual balancing that Stilz proposes strongly favors the
rights of the latter group over those of the former. Indeed, the argument is
not so much one in which “balancing” is central, but rather one in which
prescription or supersession of the rights of displaced persons takes center
stage. “Wrongs of displacement” are superseded where the new residents of
seized territories meet Stilz’s conditions for having occupancy rights and
where the displaced persons are not left stateless (or as second-class citizens
in some other state). If displaced persons fail to secure equal citizenship
elsewhere, they retain a “right of return” to the land from which they were
wrongly displaced, but no “right to expel” the current residents.45
Stilz defends her position on rights-supersession primarily by citing
Jeremy Waldron’s account of the matter. But Waldron’s account (like all
such accounts) is most forceful with respect to long-term supersession, where
the issue is one of ancient wrongs, where the cast of characters is now com-
pletely changed, where facts about the wrongs are obscure, and where judg-
ments about “what would have happened” in the absence of the wrong are
very tenuous. In cases of that sort, it is, indeed, difficult to determine the
structure of present-day rights by appealing to facts about historical wrong-
ings of the original rightholders. Stilz, however, is offering us instead an
account of occupancy rights according to which they can be acquired very
rapidly indeed (in “the second or third generation”46) by some of the bene-
ficiaries of wrongful territorial seizures.
The property principle that would be the analogue of Stilz’s occupancy
principle would work as follows. If you at gunpoint force me and my family
44
Stilz (2011), 584.
45
Stilz (2011), 586.
46
Stilz (2011), 585. It is not immediately obvious why even the “fault” of the original
aggressive displacers should not also be capable of being “superseded” on an account like
Stilz’s. Her acceptance of the supersession of the victims’ rights seems to grow almost
entirely out of her concern that once the new occupants are “settled” in the territory, it
will be “impossible to move [them] without damage to nearly all [their] life plans” (584);
this is “what really counts for supersession” (583). But if wrongful aggressors manage to
refocus their life plans quickly enough, shouldn’t they also acquire rights of occupancy
(that supersede those of displaced parties) in this way? Stilz has more recently added the
further requirement that one can acquire an occupancy right only if “his connection to
the territory was established without any wrongdoing on his part” (Stilz [2013], 354). But
if historical wrongdoing does, indeed, in this way limit the acquisition of occupancy
rights, it is unclear why the wrongdoers’ second- or third-generation descendants—who
typically enjoy the ill-gotten territory in full (or at least adequate) knowledge of the moral
impropriety that produced their enjoyment—do not also count as complicit in the
wrongdoing (like the knowing recipients of stolen goods).
168 A. John Simmons
from our family home and then succeed in raising your children and grand-
children in it (while my displaced family manages acceptably in the new
place to which we’ve fled), your family acquires an uncontested right to my
home. Indeed, even if my family doesn’t manage acceptably in our new loca-
tion, we retain at most a right to come back and share the house with your
family. This to me has less the look of a reasonable “balancing” of compet-
ing legitimate claims to the home than it does the look of a disturbingly
quick derivation of right from might. While Stilz assures us that her account
of territorial rights is not “a mere legitimation of the status quo,”47 it must
be allowed that a very strong conservative bias colors at least her treatment
of the “defeater” claims involved in cases of wrongful displacement.
Consider now Stilz’s account of wrongs of annexation. Because wrongful
annexation is a matter of violating a people’s right, according to Stilz, it fol-
lows that there is nothing obviously wrong with states annexing as much
territory as they please, provided that the territory in question is occupied
only by politically unorganized groups or individuals (or by [non-people]
subjects of some illegitimate state). While Stilz says that annexation must
still be “justified” even in such cases,48 it is very unclear how a legitimate
state could fail to be justified in such annexations—given the basic Kantian
(coercively enforceable) duty on each person to submit herself to legitimate
institutions for the administration of justice. But this appears to mean that
any legitimate state possesses a more-or-less permanent right to annex any
territory that is occupied only by persons who are not citizens in another
legitimate state—even by, say, that land’s harmless original occupants—and
this despite the fact that such persons may constitute no threat at all to
anyone in any legitimate state, at most “threatening” only those who freely
choose to share their lifestyle there.
Indeed, given that such persons do not enjoy “legal residence,” existing
legitimate states would appear to be entitled not merely to annex their land
but also to expel the land’s inhabitants. Stilz plainly wants to deny this,
affirming such inhabitants’ rights of occupancy; for she now characterizes
individuals’ occupancy rights as “preinstitutional” and uses as a central illus-
tration the case of the Navajo tribe, who “were not a politically organized
group.”49 But notice how she defines this preinstitutional occupancy right:
“the right of individuals to live in a certain area, and, together with others,
to authorize a legal institution to enforce rules regarding ownership, or to
engage in social practices defining their ownership.”50 Here, Stilz appears
47
Stilz (2011), 599–600 (n 42).
48
Stilz (2011), 598 (n 39).
49
Stilz (2013), 324 (n 1) 333.
50
Stilz (2011), 333.
to begin parting company with Kant; for with mere “social practices” gov-
erning relations of ownership—rather than with a definite political/legal
institutional structure doing the job—the resulting property rights may
well lack the determinacy, stability, and enforceability required for them
to be (in Kant) “conclusive” moral rights. More important, though, those
individuals who are not engaged with others in collectively defining owner-
ship appear on this account to be denied preinstitutional occupancy rights,
making them morally vulnerable, if not to expulsion, then at least to subjec-
tion (through annexation of their land) by any legitimate state. But surely
even in such cases there is something wrong with simply taking or with
subjecting to outside control land that is being innocently and productively
used by individuals, at least where there is plenty of land left for others to
use as they please. The wrong in question here may involve neither wrongful
displacement of individuals (since such persons may be subjected to state
authority rather than displaced) nor wrongful annexation of a people (since
such persons lack a “people’s” political/legal organization). It is, instead, a
simple case of (what we can call) “wrongful subjection.”
As examples of such wrongful subjection, consider first an individual,
living in isolation from other persons (such as an idealized version of
Thoreau), who develops a conception of the good that is both firmly
opposed to membership in any political organization and revolves crucially
around his relationship to the particular land on which he resides. Consider,
next, a loose group of individuals (such as American frontier settlers, living
on widely separated bits of land) who mostly just mind their own business,
but who also together develop the same kind of ruggedly independent,
anti-political, territory-specific conception of the good. Finally, imagine a
socially well-integrated group—but one that lacks the formal, rule-centered
structure necessary for legitimate government and law (being, e.g., both
anti-democratic and “impressionistic” with respect to property and basic
rights)—that also has an anti-political and territory-specific conception of
the good (such as the Seminole “nation,” prior to its expulsion from Florida).
Now suppose that these individuals and groups are forcibly subjected by
others to (structurally just) institutions of government and law, making a
good life for them impossible.
Stilz appears to be committed to denying that these acts of forcible sub-
jection are wrongs (except, perhaps, in the case of the Seminole, depending
on how we understand their social practices regarding property). Her rea-
son, I suppose, must be that where people live outside of legitimate states
(or legitimating “social practices”), they live in a condition of injustice with
respect to those around them. If legitimate states subject (or expel?) such
persons, their doing so is defensible as an act required by justice. But if that
is indeed her answer, surely it misses the most important part of what
170 A. John Simmons
makes the coercive subjection or displacement of innocents wrong. While

sometimes doing so may rob the expelled of a “stable legal residence” that is
central to their lives, the deeper wrong is surely that it simply wrecks their
lives by robbing them of a territorially located and firmly rooted way of life.
Territorially grounded conceptions of the good structure and help to give
meaning to the lives they guide, and they do this regardless of whether those
lives involve acting with others to “authorize legal institutions” or to engage
in “social practices” that define ownership. Peoples are not all that matters
morally. People matter too. Individuals’ interests in the land on which they
live and work can be of the same kind and urgency whether or not they are
socially engaged with others in “defining ownership;” in either case, to
employ Stilz’s words, “occupancy of a particular place is of central impor-
tance for [the] individual’s life-plans and projects.”51
In Stilz’s treatment of annexation, then, just as in her treatment of displace-
ment, she seems to take only a half-step towards accepting the moral relevance
of the history of states’ acquisitions of territory. In both of the “defeater” cases,
individuals and groups with strong and innocent-looking connections to par-
ticular portions of the earth are denied (by Stilz’s account) rights to control
over their land or rights (when wrongfully displaced) to regain that land. But
worse, even such half measures appear to sit very uncomfortably with the
Kantian foundations of Stilz’s position. As we have seen, the core of her posi-
tion is a Kantian functionalism: states’ legitimate authority over us and our
duties to uphold our states derive from the state’s performing its mandatory
function of making justice possible. How we happen to find ourselves subject
to some legitimate state’s justice-administering institutions seems simply
beside the point. “If a state exists and enforces a legitimate system of property
law, it necessarily represents me,” Stilz says, “we have a duty to support and
comply with legitimate institutions where they exist” (and, we might add [to
complete the Rawlsian thought] where “they apply to us”).52
Nowhere in this basic line of argument is there any obvious motivation
for concern about how we came to be subject to a just state (or for the
“fault” that might be ascribed to those who do the subjecting), and Stilz
dismisses as “misguided” voluntarist worries that we (typically) have no
choice in our inclusions in our various political collectives.53 Yet, in her
treatment of the two defeater claims, she attempts to take seriously some of
the non-functionalist, historical considerations that are crucial to any ade-
quate understanding of states’ claims to legitimate authority. And there,
perhaps surprisingly, it turns out that even “an absolutely perfect state,” one
51
Stilz (2011), 334.
52
Stilz (2011), 581, 582.
53
Stilz (2011), 597.
that performs its mandatory justice-related functions as well as a state could,

nonetheless lacks territorial rights if it has acquired its control over territory
in the wrong way.54 History does matter—but apparently only sometimes
and only for some people.
My deeper worry about Stilz’s position, then, is that its historical side
cannot be adequately motivated by its core functionalism, but operates sim-
ply as an ad hoc addendum to it (designed to try to avoid the counterintu-
itive implications of that position). If I am right in this, then Stilz’s own
theory is an “unstable hybrid,” like the plebiscitary voluntarism she rejects
for this reason.55 My suspicion is that the same will prove true of all other
efforts to avoid the counterintuitive implications (outlined above) of the
other, non-functionalist alternatives to Lockean voluntarism. Pluralist theo
ries are not, of course, indefensible simply in virtue of their pluralism. But
defensible pluralism requires a unifying overall motivation, not simply a
convenient conjunction of elements deriving from views that are essentially
in tension with one another. The Lockean theory I favor is itself pluralistic,
since it limits individuals’ historical claims to land and resources (and thus
also limits legitimate states’ claims to territory) with the requirement that
each person (and, so, each person’s state) take no more than a fair share of
the earth (to which no person naturally has better claim than any other).
But this pluralism is taken seriously from the start in the style of Lockean
theory that I support. Other hybrid views, like Stilz’s, appear to define a
“default” position on state territorial rights first, appealing only to the core
orientation of the theory (e.g., pure functionalism). Only later, in the face
of concerns about the pure theory’s disturbing implications are other con-
siderations (e.g., about historical wrongs) allowed to have “defeating” moral
weight within the theory, thus creating the impression of merely convenient
conjunction. A defensible hybrid theory must be pluralistic from the start,
driven by a compelling unifying motivation and defending an ideal of right-
ful territorial claims that takes seriously throughout all of the factors that
help to define that ideal. Lockean voluntarism, I believe, unlike the rival
theories of state territorial rights I’ve considered here, satisfies that most
central theoretical requirement.
Bibliography
Altman, Andrew and Wellman, C. H. (2009). A Liberal Theory of International
Justice (Oxford: Oxford University Press).
54
Stilz (2011), 583.
55
Stilz (2011), 600.
172 A. John Simmons
Buchanan, Allen (2003). “The Making and Unmaking of Boundaries: What

Liberalism Has to Say,” in A. Buchanan and M. Moore (eds), States, Nations and
Borders (Cambridge: Cambridge University Press), 231–61.
Buchanan, Allen (2004). Justice, Legitimacy, and Self-Determination (Oxford: Oxford
University Press).
Christiano, Thomas (2006). “A Democratic Theory of Territory and Some Puzzles
about Global Democracy,” Journal of Social Philosophy 37:1, 81–107.
Dworkin, Ronald (2011). Justice for Hedgehogs (Cambridge, MA: Harvard University
Press).
Gans, Chaim (2003). The Limits of Nationalism (Cambridge: Cambridge University
Press).
Meisels, Tamar (2009). Territorial Rights, 2nd edn. (Dordrecht: Springer).
Miller, David (2007). National Responsibility and Global Justice (Oxford: Oxford
University Press).
Nine, Cara (2012). Global Justice and Territory (Oxford: Oxford University Press).
Pogge, Thomas (1992). “Cosmopolitanism and Sovereignty,” Ethics 103:1, 48–75.
Pogge, Thomas (1994). “An Egalitarian Law of Peoples,” Philosophy & Public Affairs
23:3, 195–224.
Rawls, John (1999). The Law of Peoples (Cambridge, MA: Harvard University Press).
Sidgwick, Henry (1908). The Elements of Politics (London: Macmillan).
Simmons, A. John (2001). “On The Territorial Rights of States,” Philosophical Issues
11: Social, Political, and Legal Philosophy, 300–26.
Simmons, A. John (2010a). “Ideal and Nonideal Theory,” Philosophy & Public Affairs
38:1, 5–36.
Simmons, A. John (2010b). “Disobedience and Its Objects,” Boston University Law
Review 90:4, 1805–31.
Simmons, A. John (2013). “Democratic Authority and the Boundary Problem,” Ratio
Juris 26:3, 325–56.
Steiner, Hillel (2005). “Territorial Justice and Global Redistribution,” in G. Brock
and H. Brighouse (eds), The Political Philosophy of Cosmopolitanism (Cambridge:
Cambridge University Press), 28–38.
Stilz, Anna (2009). “Why Do States Have Territorial Rights?,” International Theory
1:2, 185–213.
Stilz, Anna (2011). “Nations, States, and Territory,” Ethics 121:3, 572–601.
Stilz, Anna (2013). “Occupancy Rights and the Wrong of Removal,” Philosophy &
Public Affairs 41:4, 324–56.
Waldron, Jeremy (1992). “Superseding Historic Injustice,” Ethics 103, 4–28.
Wellman, C. H. and Cole, Phillip (2011). Debating the Ethics of Immigration
(Oxford: Oxford University Press).
7
Can Reductive Individualists Allow
Defense Against Political Aggression?
Helen Frowe
INTRODUCTION
Collectivist accounts of the ethics of war have traditionally dominated just

war theory (Kutz 2005; Walzer 1977; Zohar 1993).1 These state-based
accounts have also heavily influenced the parts of international law pertaining
to armed conflict. But over the past ten years, reductive individualism has
emerged as a powerful rival to this dominant account of the ethics of war.
Reductivists believe that the morality of war is reducible to the morality of
ordinary life. War is not a special moral sphere with its own special moral
rules. Reductivists typically reject a collectivist approach to the morality of
war in favor of an individualist view, according to which individuals (rather
than states or other collectives) are the proper focus of moral guidance and
evaluation. This view holds that the rules governing killing in war are simply
the rules governing killing between individuals, most obviously the rules of
self-defense and other-defense (Fabre 2009; Frowe 2014; McMahan 2009;
Tadros 2014).
This chapter defends reductive individualism against the claim that it is
unable to sanction wars of national defense that seek to protect non-vital
interests. These non-vital interests include political goods such as the defense
of sovereignty—that is, of political and territorial integrity.
The most detailed articulation of this claim comes from David Rodin
(Rodin 2014). Rodin’s argument has two main components, which I shall
label the Conditional Force Argument and the Proliferation Problem. The
Conditional Force Argument holds that, by the reductivist’s own lights, wars
1
Versions of this chapter were presented at an Ethics of War conference at UC San
Diego, a Pacifism conference at the University of Belgrade and the Workshop in Political
Philosophy at the University of Arizona. Many thanks to the audiences for their helpful
comments. Special thanks to Kim Ferzan for very helpful suggestions.
174 Helen Frowe
that seek to defend only political goods are necessarily disproportionate and
therefore always unjust. The Proliferation Problem holds that there is no
morally significant difference between states and some other collectives.
Some non-state collectives can embody whatever values we think warrant
lethal defense of the state (Rodin uses the example of the Cadbury chocolate
company, whose employees formed a distinctive community that was
destroyed when Cadbury was the subject of a hostile takeover by Kraft in
2010). So, even if we could show that it is proportionate for states to wage
defensive wars against threats to non-vital interests, we would lack grounds
for restricting this permission to states, which seems like a worrying result.
Rodin concludes that we should grant defensive rights to neither states
nor other collectives, and his conclusion thus results in a form of limited
pacifism. Unless a state is faced with genocidal aggression, Rodin thinks it
impermissible to resist for the state to forcefully resist invasion. It’s worth
noting, then, that Rodin is not conceiving of his argument as an objection to
reductive individualism. On the contrary: he thinks it’s true that states lack
a right to defend non-vital interests. I argue here that it’s not true, and that
reductivists aren’t committed to this claim.
In section 1, I outline the Conditional Force Argument in more detail. This
argument is underpinned by a particular account of our moral responsibility
for what I call “mediated harms.” In section 2, I argue that this particular
account of mediated harms is mistaken. Part of the problem with this
account is its reliance on a distinction between lesser interests and vital interests.
I argue in section 3 that this distinction also renders Rodin’s account of the
difference between political aggression and genocidal aggression incoherent.
In section 4, I defend a different account of our moral responsibility for
mediated harms. In section 5, I argue that the defense of aggregated lesser
interests can warrant endangering vital interests. In section 6, I consider the
Proliferation Problem, and argue that states (unlike Cadbury) protect our
right to form communities. It’s the right, and not communities, that matter
for permissible defense.
1. THE CONDITIONAL FORCE ARGUMENT
Imagine that a state wants to expand its territory by annexing land belong-
ing to another state. The aggressor state will use force only if the members
of the victim state resist the annexing. This kind of aggression—which will
become violent only if it is resisted—is what is sometimes called in the liter-
ature a bloodless invasion or, as I will call it, conditional force. Most accounts
of jus ad bellum hold that this kind of annexing of land constitutes a just cause
for defensive war on the part of the victim state. One plausible explanation
Reductive Individualism & Political Aggression 175
of why war would be a proportionate response to this sort of aggression is

that the goods directly threatened by the invasion—those of political and
territorial control—warrant protection by lethal force.
Because sovereignty is usually conceived of as a collective good, it might
look like this is the source of the difficulty for a reductive individualist. But
this would be too quick. The reductive individualist can plausibly argue that
sovereignty has merely instrumental value that comes from its ability to
enable states to facilitate and protect things that are of value to the state’s
individual members. This seems to me to be roughly the right explanation
of why sovereignty matters, and it explains this without positing the exist-
ence of some irreducibly collective good. This account of sovereignty also
supports the view that states enjoy only conditional sovereignty—they have
rights against interference insofar as they protect and meet the rights of
their individual citizens, but not independently of protecting those individ-
uals’ rights. This also seems to me correct.
But proponents of the Conditional Force Argument point out that the
mere fact that we can show that sovereignty is valuable at an individual level
does not show that it is permissible to defend this good by means of war,
where this involves intentionally lethally harming combatants and foreseea-
bly lethally harming non-combatants. Lethal harms can be permissible only
if they are proportionate to the good being defended. But by rejecting the
collective value of sovereignty and reducing its value to its importance
to individuals, the reductivist account lacks an explanation of why it
is proportionate to wage war to defend this individual good (Lazar 2014;
Rodin 2014).
For example, Rodin argues that when it comes to individual self-defense,
lethal force is a proportionate response only to threats to a person’s vital
interests (Rodin 2014: 81). An individual may kill to defend herself against
very serious harms such as death, torture, slavery, and rape. But she may not
kill to defend her lesser interests, which for Rodin includes property and
other resources, and political rights such as the right to vote. She may not,
for example, kill a person who is trying to prevent her from voting on the
day of an election, perhaps by damaging her only means of transport to the
polling station. Even though this person will violate her right to vote and to
political representation, these rights are lesser interests, insufficiently impor-
tant to warrant lethal defense. Just as importantly, she may not endanger
bystanders’ lives in order to defend her political rights.
If this is true for individuals, Rodin says, reductivists must grant that it
applies at the national level. Rodin identifies two forms of aggression cor-
responding to threats to lesser interests and vital interests. Political aggres-
sion is aggression that directly threatens only lesser interests (with a
conditional threat to vital interests if the victim state does not capitulate).
176 Helen Frowe
So, the annexing of land described above counts as an act of political aggres-
sion on this account.
Genocidal aggression is aggression that directly threatens vital interests—
that involves killing and otherwise harming the vital interests of members
of the victim state even if they do not resist. Such harming could be the end
of the aggression—it could, for example, be a war aimed at ethnic cleansing.
Or, it might be a means employed to some other end, such as access to
resources, but where no opportunity is offered for the victim state to capitu-
late. Rodin’s account somewhat stretches our ordinary usage of the term “gen-
ocidal” to include, for example, widespread raping, maiming, enslavement,
long-term displacement, and rending people stateless, but this extension
needn’t trouble us here.
Rodin argues that political aggression and genocidal aggression “have a
fundamentally different moral structure. Resistance to political aggression
will typically produce greater loss to vital interests among individuals within
the defended group, compared with engaging in no defence at all” (Rodin 2014:
87). Given this, resistance to political aggression is usually “straightfor-
wardly self-defeating and hence morally irrational” (Rodin 2014: 82). In
addition, since resistance involves risking the lives of other people for the
lesser interests of political goods, resistance is morally impermissible. If no
member of the polity may kill or endanger lives to defend her individual polit-
ical rights, then the polity as a whole may not kill or endanger lives to
defend those rights. Thus, it is only genocidal aggression that may be per-
missibly resisted, since here one endangers vital interests in order to protect
vital interests.
According to Rodin, the only way to explain why something like sover-
eignty warrants lethal protection is to attribute some kind of inherent value
to the state or political community (Rodin 2014: 69). This move is not
open to the reductive individualist, given her rejection of the idea that there
are collective values that are irreducible to their value for individuals. So, it
looks like the reductivist is committed to saying that sovereignty may not be
defended by means of war after all. War always involves inflicting lethal
harms and will therefore always be disproportionate when it is aimed at the
defense of the lesser interests of individual citizens.
2. MEDIATED HARMS
A natural response to the Conditional Force Argument is to point out that

members of a state faced with merely political aggression need not make
immediate recourse to all-out war. Surely, we might say, only full-scale mili-
tary resistance is disproportionate—non-violent resistance is not dispropor-
tionate. And should such resistance be met violently—well, then the vital
interests of the citizens are directly threatened, and forceful resistance will
be a proportionate defense of those vital interests.
But Rodin argues that this reply fails because any calculation about the
proportionality of this non-violent resistance must take into account the harms
that the victim state foresees will be inflicted by the aggressor in response
(Rodin 2014: 82). Call these harms “mediated harms.” If the victim state pre-
dicts that its initial defense will cause the aggressor to escalate the situation,
that initial defense can be disproportionate. We can see this by thinking about
cases like Pinch:
Pinch: Bully wants to pinch Victim’s arm, which will hurt. Victim knows that if (and
only if) he tries to prevent this, Bully will become so enraged that he will kill five people.
It seems plain (to both me and Rodin, at least) that Victim may not try to
prevent Bully’s pinching his arm.2 Rodin grants that since Bully (and not
Victim) will inflict these mediated harms, they don’t weigh as heavily in the
proportionality calculation as harms that Victim will inflict himself. But the
fact that the harms will arise through someone else’s intervening agency
doesn’t make them irrelevant to what Victim ought to do. While mediated
harms are somewhat “discounted”—they don’t count as much for propor-
tionality as harm that one directly inflicts oneself—they are not discounted
to zero. They still matter for determining what one is permitted to do.
If this account of our responsibility for mediated harms is correct, the
reductivist must hold that the same reasoning applies at the level of war.
When the members of a state face a threat, they cannot judge whether war
or even non-violent resistance would be a proportionate response to that
threat unless they factor in how the aggressor will likely react. If they predict
that the aggressor state will react by waging a war that endangers the lives of
many people, they must take those foreseen harms into account.
Of course, the harms that the aggressor state will inflict are discounted in
the victim state’s proportionality calculation. But, Rodin argues, when it
comes to war, this discount is effectively canceled out because of the duty of
care that the members of the victim state owe their fellow citizens, which he
illustrates with the following case:
2
We might object that to prohibit Victim’s defending himself is to incentivize wrong-
doing. There are two reasons why I don’t think this shows us that Victim may defend
himself in Pinch. The first is simply that having any kind of proportionality constraint
makes it possible for a person to threaten you under circumstances in which it is imper-
missible for you to respond. Second, there’s no reason why Victim can’t factor predicted
long-term mediated harms into his proportionality calculation. If, in the long run, it
seems likely that acquiescing to a particular threat will incentivize wrongdoing that
causes more mediated harm overall than resisting the threat, it could be proportionate for
Victim to resist.
178 Helen Frowe
House: A villain invades and occupies your home without justification. He lives in
your house and eats your food, forcing you and your family do all the work. He
makes it clear that he will not use force unless you resist. There is no end in sight.
You could tackle the invader, but if you do so, it is likely that one of your children
would be killed, either as a side effect of the struggle, or as a punishment for your
resistance. (Rodin 2014: 84–5)
Rodin argues that most people would think it impermissible to tackle the
villain, because “the value of a child’s life – especially since this is your child
and you owe a profound duty of care to him” outweighs the good of evict-
ing the occupier (Rodin 2014: 85). He thinks that a similar duty obtains
when members of a state are considering going to war. The people whom
they will be endangering through their resistance are “people who are
bound to [them] by relationships of loyalty, community and kinship.”
They “may be our comrades in arms, our family members, our neighbours”
(Rodin 2014: 83). So, any discount that arises from the fact that it is the
aggressor state that will be endangering these people is undone by the fact
that the members of the victim state owe a duty of care to the people who
are endangered.
The members of the victim state must, therefore, proceed as if they them-
selves are harming their co-citizens as a side effect of their self-defense. Most
people think that one may not, for example, kill a bystander as a side effect
of defending oneself against a broken leg. Similarly, Rodin argues that one
may not endanger one’s fellow citizens’ vital interests for the sake of defend-
ing something of lesser interest, which for Rodin includes land, resources,
and political independence. Exposing people to such harms is dispropor-
tionate compared to the goods threatened by purely political aggression.
Reductivism is, therefore, unable to allow that resistance against political
aggression can be just.
3. MEDIATED HARMS AND VITAL INTERESTS
While Rodin’s account of meditated harms seems to get the right result in
Pinch, it’s less plausible in other cases. Take Rape:
Rape: Angry Rapist tries to rape Alice. If (and only if ) Alice fends him off, he will be
so angry that he will go and rape two other women. If (and only if ) Alice kills Angry
Rapist, Angry Rapist’s friend will go and rape two other women.
It seems to me very plausible to think that Alice may defend herself against
Angry Rapist even if she foresees that her defense will trigger the rape of two
other people. And, yet, the harm she predicts will arise from her defense is
twice that which will befall Alice herself. Even if we discount the foreseen
harm by fifty percent, in virtue of the fact that it is mediated harm, Alice
would still be triggering the equivalent of one rape as a side effect of defend-
ing herself against rape.
Rodin’s view entails that Alice must proceed as if she would be inflicting
the discounted harm herself—as if she would be inflicting the rape upon the
other woman as a side effect of her defense. But most accounts of permissi-
ble defense do not allow someone to inflict as a side effect a harm that is
equal to the harm she is seeking to avoid. For example, I may not divert a
runaway trolley from where it will kill me down a sidetrack to where it will
kill some other innocent person. Rather, I would be permitted to divert the
trolley only if the person on the sidetrack would suffer a significantly lesser
harm than the harm facing me. If this is right, even if we discount the medi-
ated harm by fifty percent, Rodin’s account still prohibits Alice’s defending
herself in Rape.
This seems like the wrong result to me. Rodin might reply that mediated
harms can be more heavily discounted than I have suggested here. As I will
argue in section 4, I think they are much more heavily discounted. But this
isn’t going to help Rodin’s account, because this view requires that however
much we discount the mediated harms, the duty of care that we owe to our
co-citizens cancels out this discount. This means that Alice may not fend off
Angry Rapist when doing so will cause him to rape or seriously assault just
one other woman if that the other woman is a co-citizen.
Of course, Rape illustrates the defense of a vital interest—the right not to
be raped. We might think that we should therefore treat Rape differently to
Pinch, which involves the defense of a lesser interest—the right not to be
pinched. We might say that while mediated harms can make defense of lesser
interests disproportionate, they can’t make the defense of vital interests dis-
proportionate. We may always defend our vital interests even if we foresee
mediated harms to other innocent people’s vital interests.
But this response is pretty ad hoc: it’s not clear why our account of
responsibility for mediated harms should be sensitive to this distinction
between lesser and vital interests. As Rodin says, “[d]efensive action is
impermissible when it foreseeably produces harmful effects that are dispro-
portionate to the good one is seeking to achieve” (Rodin 2014: 77). This claim
cannot plausibly be restricted to the defense of lesser interests—it’s simply
true of defensive action in general, which includes the defense of vital inter-
ests. If mediated harms count among our harmful effects—as Rodin and
I think they do—they must be able to make defensive action disproportion-
ate even when one is defending vital interests.
Moreover, it’s not clear that this response is even open to a proponent of
the Conditional Force Argument. Given, for example, Rodin’s claim that
“[t]he morality of resisting the direct threat component of a conditional
180 Helen Frowe
attack cannot be determined independently of the conditional threat,” we

cannot simply stipulate in advance that vital interests always or even usually
warrant defense by lethal force, irrespective of the consequences (Rodin 2014:
82). Whether or not they do so will depend on what the aggressor has
threatened to do if the vital interest is defended: the central claim of the
Conditional Force Argument is that the proportionality of defense is par-
tially dependent upon the predicted mediated harms.
4. GENOCIDAL AGGRESSION
This line of argument also reveals two difficulties with the distinction between
political aggression and genocidal aggression that corresponds to the distinc-
tion between threats to lesser interests and threats to vital interests. The first
concerns Rodin’s claim that his account allows that “coordinated, forceful
defence against genocidal aggression can be morally justified in reductivist
terms” (Rodin 2014: 83). We’ve just seen how Rodin’s account makes it
impermissible for Alice to defend her vital interests if, in doing so, she will
expose others to similar harms. As we saw, it could be the case that, even
discounted, the mediated harms are similar in weight to the harms facing
Alice, and thus she would not be permitted to defend herself while foresee-
ing those harms to others.
If so, it must also be true—if we are reductivists—that states may not
defend their vital interests when doing so exposes others to similar harms (as
defensive war inevitably does) even once we apply the discount. For exam-
ple, a state with a small population facing a genocidal threat may foresee
that resisting will result in a war that will cause lethal and other very serious
harms to a large number of people: perhaps more people than whose lives
are threatened by the genocide. If resistance risks the vital interests of twice
as many people as are threatened by the genocide, even discounting these
mediated harms by fifty percent will not make resistance permissible, just as
it was not permissible for Alice to resist if she foresees that the equivalent of
one other rape will be inflicted on another innocent person. In other words,
Rodin’s account of mediated harms will sometimes prohibit resisting even
genocidal aggression.
The second problem is that there will be cases in which Rodin’s account sim-
ply fails to classify an act of aggression as either political or genocidal, and
thereby fails to tell us whether such aggression may be resisted. Consider Threat:
Threat: An aggressive state is demanding the execution of all 1000 members of a
small ethnic minority group living in the victim state. If the victim state does not
kill these citizens or permit a delegation from the aggressor state to do so, the
aggressor will invade. During the invasion, 100 members of the minority group will
be killed, along with 1100 defending combatants and innocent non-combatants in
the victim state.
So, a total of 1200 innocent citizens will suffer lethal harm if the victim state
decides not to capitulate to the aggressor’s demands to execute the 1000. Must
the victim state capitulate?
What Rodin is going to say about Threat will depend on how he is identify-
ing the relevant group, harms to which we must factor into our proportion-
ality calculation. His definitions of political aggression and genocidal
aggression are tied to thoughts about the harms that will befall the group being
defended—he speaks of “those very people whom the defensive action is
intended to defend ” and “the defended group” (Rodin 2014: 87). The ques-
tion, then, is whether these references to people and groups are meant to
pick out all members of the victim state that is being conditionally threat-
ened, or just the sub-group that is being directly threatened (in this case
the members of the ethnic minority).
If the “group being defended” is meant to be synonymous with “all mem-
bers of the victim state,” that will give us the result that the aggression in
Threat is political aggression by Rodin’s lights. Fewer vital interests will be
harmed overall within this group if the victim state capitulates and executes
the members of the ethnic minority. Resistance against this aggression will
thus be impermissible. This result seems to me wrong in both its denial that
the threat is genocidal, and in its seeming pronouncement that the victim
state ought to execute the members of the ethnic minority group rather
than defend them.
So, I should think that Rodin will want to go the other way, and argue that
references to “the group being defended” are not meant to be synonymous
with “all members of the victim state,” but can rather refer to a sub-group
within that state, such as the ethnic minority group I have described. He
tells us that an aggression is political, and should not be resisted, if resisting
it causes more harm to the vital interests of “those very people whom the defen-
sive action is intended to defend.” Since in my example, resisting will be less
harmful to the vital interests of the minority group whom the action is
intended to defend, Rodin can count the threat as a genocidal aggression.
But this result isn’t just semantic in his view—it’s normative. It tells us
that the invasion may be resisted, because it’s genocidal. And that seems
right. But this is a case in which more harm will befall the vital interests of
innocent people overall if the invasion is resisted, which was meant to be the
definition of political aggression and part of the explanation of why resist-
ing political aggression is wrong: resisting in such circumstances is, accord-
ing to Rodin, morally irrational. So we have two problems. One is that this
182 Helen Frowe
case seems to defy categorization on Rodin’s account. This problem comes

from his reliance on lesser and vital interests to classify forms of aggression,
and I return to it below. The second problem is that it looks like the only way
that Rodin can explain why it’s permissible to resist the aggression in Threat is
if he ignores the harms that will befall people within the victim state who are
not members of the minority group. But then we need an argument for why
the interests of those people don’t matter for the purposes of judging whether
the resistance is permissible.3 I’m not sure what that argument would look
like. They are all co-citizens, so any discount attached to the mediated harms
will be canceled out by the alleged duty of care. The government must thus
proceed as if they are themselves inflicting these additional deaths, which
looks like disproportionate, and therefore impermissible, defense.
I think that the correct answer in Threat is that it the threat is genocidal,
and that it may be resisted even if the minority group’s defenders foresee that
their resistance will cause the aggressor to wage a war that costs more lives
overall, similarly, Alice may resist in Rape even if Angry Rapist (or his friend)
will then rape other people. So, the reductivist needs an account of our
responsibility for mediated harms that explains this permissibility.
5. MEDIATED HARMS AND RESCUE CASES
I think that our judgments about how much harm Victim ought to bear to
rescue people from harm can usefully inform our judgments about how
much cost Victim must bear in a mediated harm case rather than endanger
some other innocent person.
The connection between the two sorts of case might be explained by
Victor Tadros’s suggestion that both the restrictions on harming people as a
means for the greater good and the restrictions on requiring people to bear
costs to rescue other people “have as their source our status as ends. Because
we are not available as tools for the sake of each other, we cannot be harmed
as a means to a greater good and duty does not require us to act for the greater
good” (Tadros 2011: 250). If Alice were not permitted to defend herself against
Angry Rapist, we would effectively be requiring Alice to treat herself as a
means, permitting Angry Rapist to harm her for the sake of others. We can
sometimes be so required, but there are limits to how much cost we can be
required to bear to such ends, just as (and for the same reason as) there are
limits on the costs we must bear to rescue others.
3
Even if we accept Rodin’s duty of care argument, which could explain why we need
not care too much about harms to non-combatants in the aggressor state, this will not
explain why the vital interests of other co-citizens can be ignored in our calculations
about whether resistance is proportionate.
Compare River with Bomb:

River: Five people are drowning on a life-raft in dangerous waters. Runner can pull
the life-raft to shore, but will suffer the loss of his leg in doing so.
Bomb: Attacker attacks Victim. If (and only if ) Victim fends Attacker off, Attacker
will detonate a bomb that will kill five people. If Victim doesn’t fend Attacker off,
Attacker will cut off his leg.
Let’s stipulate that for any cost lower than the loss of his leg, Runner would
be required to rescue the five on the life-raft. Since, in River, he can rescue
them only at the cost of his leg, he isn’t required to rescue them.
I also think that Victim is not required to suffer the loss of his leg in
Bomb, even though I think that we are typically more responsible for medi-
ated harms than for harms we fail to prevent. This is because the kinds
of mediated harm case we’re discussing—defense cases—involve a further
cost to Victim that isn’t usually present in a rescue case. In River, even if
Runner bears the cost of the loss of his leg, it doesn’t look as if his rights
have been violated. Either saving the five is supererogatory, in which case
Runner has chosen to sacrifice his leg, or it’s obligatory. It’s hard to see how
in this case doing what’s obligatory could also constitute a rights-violation.
In Bomb, in contrast, Victim must suffer not only the loss of his leg if he
refrains from defense, but also a serious rights-violation at the hands of
Attacker. This additional cost—the fact that losing his leg will violate his
rights—could counteract his greater responsibility for the harms in Bomb
compared with Runner’s responsibility for the harms in River. The upshot
would be that, in defense cases, Victim is required to bear roughly the same
amount of cost to prevent mediated harms to others as he would be to
rescue others from harm.
This will set the threshold for proportionate defense much lower than
the fifty percent I hypothesized earlier. Mediated harms will be heavily dis-
counted compared with harms that we directly inflict as a side effect of
defense. Imagine that a villain ties me to some trolley tracks and sets a trol-
ley towards me. I may not divert the trolley from where it will cut off my leg
to where it will kill you. But I may divert the trolley to prevent it from
cutting off my leg even if I foresee that the villain will then lethally push you
off a cliff by way of venting his frustration at my escape.
6. A DUTY OF CARE?
This result is more permissive than Rodin’s view. Rodin denies that there is
this sort of ratio between foreseen mediated harms and the defended good,
holding simply that whenever resistance will harm a greater number of vital
184 Helen Frowe
interests among members of the victim state, the aggression is political and
the state must capitulate. This is because of his view that the discounting of
harms caused by an aggressor state in war is neutralized by the duties of care
that one owes to one’s co-citizens. One must proceed as if one were directly
inflicting the harms upon those co-citizens.
Let’s assume for a moment that Rodin is correct that there is a duty of care
that requires us to treat our co-citizens with special care. If there is such a
duty, it will apply to all co-citizens. This means that one will have a duty of
care not only to those whom one foresees will be harmed by the wrongdoing
of others, but also to those whom one can protect by resisting an invasion.
So, we will need to think about how this duty plays out in a case like Threat,
where one must choose between failing to protect a smaller number of citi-
zens from harm and causing mediated harms to a larger number of citizens.
Rodin argues that one may not intentionally kill an innocent person to
save the life of another person to whom one owes a duty of care. I may not
throw someone else’s child in front of a trolley to save my own child’s life.
But the harms to the larger number of citizens in Threat—and more gener-
ally in defensive war—are not intentional, but merely foreseen mediated
harms. We might think that this makes them easier to justify. However,
Rodin’s House case that I described in section 1 is meant to illustrate that
even when one merely foresees a mediated harm to another person’s vital
interest, one cannot bring that harm about when one owes a duty of care to
the person who will be harmed.
But this case seems to me unpersuasive, not least because it should count
as a case of genocidal aggression on Rodin’s account, and he claims that
genocidal aggression may be resisted. Rodin lists enslavement among the
infringements of liberty sufficiently serious to count as genocide. One would
also think that indefinite imprisonment counts as a sufficiently serious
infringement. To preserve the analogy between House and war, it must be
the case that in addition to being made to “do all the work,” you and your
family cannot leave the house (and even if you can, Rodin counts indefinite
displacement from one’s home as a genocidal threat when it happens at the
national level) (Rodin 2014: 79; Rodin 2014: 75).
Given this, House is a case in which more people will suffer setbacks to
their vital interests if the occupation is not resisted. Even if it is certain that
one child will die in the resistance, not resisting entails that the whole family
will be indefinitely imprisoned and enslaved. It is, by Rodin’s own lights, a
case of genocidal aggression.
So, may you resist? If Rodin’s claims about genocidal aggression are true,
then he must change his mind and say that you may—given the greater set-
back to vital interests overall, this is not a case in which it would be morally
irrational or impermissible to resist. More harm to vital interests befalls the
very group of people being defended if the occupation is not resisted, and
all members of the family are owed a duty of care.
If Rodin sticks to his guns and says that you may not resist, this implies
a ranking of interests even within the category of vital interests, according
to which it is always impermissible to expose someone to the top-ranked
harm—death—even if a greater number of other vital interests (the rights
not to be enslaved or imprisoned) are thereby harmed. But as I pointed out
above, a central premise of the Conditional Force Argument is that one
cannot stipulate the proportionality (and impermissibility) of resisting a
direct threat without taking into account the foreseen harms. So, Rodin
cannot stipulate that death always outranks harm to other vital interests,
such that it is always disproportionate to foreseeably bring about some-
one’s death while defending against a threat to less serious (but still vital)
interests. Such a stipulation would anyway be implausible. Foreseeably
causing one person’s death as a mediated harm does not seem dispropor-
tionate to freeing a larger number of people from indefinite imprisonment
and servitude.
I think it is possible that familial ties and ties of friendship can sometimes
make a difference to what we are allowed to do, although it’s not clear to me
that this is best explained as owing a duty of care to these people. But if I do
owe a duty of care as generally as Rodin supposes—essentially, to all those
people who participate in my state—it seems to me unlikely that I owe it to
them in virtue of the fact that they participate in my state. Such a duty must
rather be grounded in more general properties shared by all persons. That
this is simply a general duty owed to all persons explains why, for example,
I may not drive more recklessly in a popular tourist area, or in a foreign
country, than I drive at home.4 Such reckless endangerment could not be
justified by appeal to the absence of a duty of care, or by the claim that any
such duty is less stringent when it comes to foreign people. Nor are my
duties to rescue less stringent when I’m abroad, where all the imperiled
people are foreign. Nor does the fact that one of two drowning people hap-
pens to be British give me a reason to save her rather than the other person.
If there is a duty of care that explains why I must care about all these harms,
I should think that duty applies equally to foreigners and co-citizens. If I
can divert a runaway trolley away from where it will harm five to either:
(a) where I foresee that it will harm Local Person, or (b) to where it will
harm Foreign Person, I don’t think a duty of care requires me to choose (b)
and harm Foreign Person rather than Local Person.
4
Rodin does say that the duty of care is less important when it comes to intentionally
harming. But reckless endangerment is not intentional harming, and the tort law duty of
care is generally thought to cover just such endangerment.
186 Helen Frowe
But once this alleged duty is so widespread, covering persons in general, it

doesn’t make much sense to describe it as an extra aspect of morality that
must be factored in to our deliberations, changing what morality would
ordinarily permit, prohibit, or require. Invoking such a general requirement
to treat all people in a certain way is simply to recognize the requirements
of morality, not to discover an additional constraint on top of what morality
usually requires. If the demands that such a duty makes generalize across all
people, there is nothing in particular that the duty picks out: it does not
single out some people for special treatment. The best that Rodin can do,
absent the duty of care, is simply stipulate that morality doesn’t allow us to
expose anyone to mediated harms (to their vital interests at least). But that
is not much of an argument, and is anyway in tension with his own claims
about discounting—if it is never permissible to foreseeably bring about
such harms, what would be the point of saying that such harms are dis-
counted in our proportionality calculations?
Once we reject the duty of care argument, there will be nothing to cancel
out the discounting of mediated harms in a case like Threat. If so, it could
be permissible to protect the smaller number of vital interests by defending
the minority group in such a case, despite foreseeing mediated harms to the
vital interests of a greater number of people. The state could be required to
capitulate when the number of foreseen harms is very much greater than the
number of interests being protected. But that strikes me as a quite plausible,
rather than worrying, result for the reductivist.
If foreseen, but mediated, harms are heavily discounted in a defender’s
proportionality calculation and this discount is not canceled out, we have
made significant progress in explaining why it can be permissible for a victim
state to defend itself against both genocidal aggression in cases like Threat,
and political aggression that is only conditionally violent. Even if members
of the victim state foresee that their resistance will cause the aggressor state to
wage a war that endangers the lives of their co-citizens and of innocent
non-combatants in the aggressor state, the members of the victim state need
not proceed as if they are themselves inflicting these harms. Moreover, since
those co-citizens’ political interests are also endangered by the aggression, the
members of the victim state are acting not only in self-defense, but also in
other-defense, which has a bearing on the amount of foreseen harm it is
proportionate to bring about.
On this account of mediated harms, it would be permissible for the mem-
bers of the victim state to defend themselves against invasion unless they pre-
dicted that doing so would result in sufficiently graver harm to a sufficient
number of innocent people. The harms are sufficiently graver if the people in
the victim state would be required to save the sufferers from those harms even
at the cost of harms to themselves equal to an invasion.
7. LESSER INTERESTS AND AGGREGATION
This brings us to the crux of defense against political aggression: the jeopardiz-
ing of some people’s vital interests to defend others against threats to their
lesser interests. In making his binary division between vital and lesser inter-
ests, Rodin assumes that no number of threats to lesser interests could out-
weigh the harming of a vital interest. I think we should reject the division
between lesser and vital interests. There are some interests that will never
warrant lethal defense—there is probably no number of scratched fingers
that could make it permissible for me to avert those scratches by endangering
an innocent person’s life. But, as I will argue, some of the things that Rodin
counts as lesser interests are going to warrant lethal defense when they are
aggregated across a sufficient number of people, and also warrant triggering
mediated harms to the vital interests of innocent people. In other words,
their defense will justify both intentional lethal harms to combatants and
foreseen but unintended lethal harms to non-combatants.
Here’s an example of how so-called lesser interests can aggregate to make
lethal force proportionate. Imagine that you are going to break my arm. I
don’t think that I may kill you to stop you from breaking my arm. But if
you’re going to break lots of people’s arms, it seems to me that there will
come a point at which these harms can aggregate to make it proportionate
to kill you to prevent them (of course, other conditions such as necessity
would also need to be satisfied for killing you to be overall permissible). This
aggregation can be both interpersonal and intrapersonal. If every time my
arm heals, you break it again, I think lethal force could become proportion-
ate to stop you from continuing to break my arm. I think this is also true
when it comes to my political rights. As Rodin says, I may not kill you to
stop you from preventing me from voting. But if you’re going to stop every
member of my ethnic group from voting in an election where our impor-
tant political interests are at stake, it might well be proportionate for us to
kill you so that we can vote. And if killing you is necessary to avert this
wrongdoing, it might well be all-things-considered permissible to kill you.
It might seem as if invoking aggregation is an illegitimate move for the
reductive individualist to make, since it seems to go beyond what can be
justified in terms of individual rights. But that would be to misunderstand
the reductive individualist view. The reductivist part of the account requires
us to understand the morality of war in terms of the morality of ordinary
life. Aggregation is a familiar feature of ordinary life, and so to think it rel-
evant to war is not inconsistent with reductivism. The individualist part of
the account holds that individuals are the source of moral value and the
proper focus of moral prescription and evaluation. But this too is compati-
188 Helen Frowe
ble with caring about the number of individuals who are threatened with a
violation of their rights. All that matters for consistency with individualism
is that, ultimately, it is the individuals themselves with whom we are con-
cerned, and not with some irreducible relationship between those individu-
als. The claim that I am defending here is that a conditional invasion threatens
important rights of individuals, not of collectives.
Numbers are significant when it comes to political rights not merely for
straightforward aggregative reasons, but because the increasing numbers
can produce a step-change in what threatens each individual.5 If you dam-
age my car and prevent me from voting, this won’t really make any differ-
ence to the sort of country I live in. But imagine that measures are taken
before a UK election to ensure that anyone living outside of London is
unable to vote. This would do much more than frustrate each individual
non-Londoner’s right to vote—it would not be simply a case of iterating the
same harm of vote prevention across many individuals. Rather, the increas-
ing numbers would effect a change in the sort of harm being perpetrated
against each of those individuals, because it would change the UK from a
legitimate democratic state to an illegitimate dictatorship. States coercively
impose legal frameworks upon their citizens. A government that is elected
only by Londoners, but then attempts to rule over the whole of the UK, will
be imposing this framework illegally, unjustly coercing individuals to adhere
to it. This is a much more serious wronging of each individual citizen than
the wrong of preventing any one individual from voting, and it is this sort
of wronging that is threatened by a political invasion. An aggressor state
that seeks, for example, to replace a democratically elected government, and
threatens to meet resistance with violence, significantly worsens the condi-
tions of the citizens’ lives irrespective of what else the aggressor does.
8. THE PROLIFERATION PROBLEM
I’ve argued that Rodin is wrong about the impermissibility of resisting con-
ditional threats. I’ve done this in a way that doesn’t depend upon attributing
irreducible moral value to the state. But reductivists also need to say some-
thing about the Proliferation Problem. The challenge is not exactly Rodin’s
suggestion that other non-state groups can manifest the inherent value found
in the state. Rather, it’s that if (as I believe) there’s nothing uniquely and/or
inherently valuable about states, we want to know whether other groups of
individuals also have rights to employ lethal defense against the equivalent
of “political” threats to their integrity. To use Rodin’s example, we want to
5
Thanks to Jimmy Lenman for helpful discussion of this point.
know whether Cadbury’s employees had a right to forcibly prevent the

Kraft takeover.
I think Rodin is probably right that there is no bright line between states
and other groups when it comes to deciding which entities warrant forceful
defense against political aggression. Secessionist movements and civil wars
suggest that non-state groups are sometimes permitted to behave as states
do, defending themselves not merely against threats to their members’ lives,
but against other types of threat to members of their group. For example,
Allen Buchanan suggests that escaping the unjust redistribution of wealth
can be a just cause for secession. If a government persistently exploits the
members of some regional ethnic or national group by taxing them heavily
and yet spending disproportionately low amounts in their area, Buchanan
argues that the group can be justified in seceding from the “parent” state
(Buchanan 1998). If conditions in the parent state are sufficiently unjust
that succession is warranted, it seems plausible that the seceding group may
use force if doing so is necessary. It thus seems plausible that it’s not only
states that are permitted to use lethal force in defense of non-vital inter-
ests—states are not unique in this respect.
But I also agree with Rodin that Cadbury’s employees were not permitted
to use force to prevent the Kraft takeover. This makes it sound as if we need
some account of which sorts of groups are eligible for these defensive rights.
But I don’t think that’s the right question. The right question is which sorts
of ends justify the use of lethal force. And, as I’ve argued above, whether a
given end justifies lethal defense is going to depend on the sort of individual
rights that are at risk of being violated and the number of individuals who
rights are so imperiled.
I think there are two related points that the reductivist can make in
reply to the Proliferation Problem. The first concerns the difference between
interests and rights. The second concerns the distinctive role of the state
(and some other groups) not in protecting or preserving particular com-
munities, but in protecting their members’ rights to try to form such
communities.
9. INTERESTS AND RIGHTS
Thus far, we’ve been talking about threats to interests that are, uncontrover-
sially, protected by rights. We have rights not to be killed, be raped, be
imprisoned, be prevented from voting, have our arms broken, and so on.
We’ve assumed throughout that the victims of threats to their interests are
innocent people who have done nothing to forfeit their rights that protect
those interests. Even if Rodin and I disagree about whether one may violently
190 Helen Frowe
defend one’s right to vote, we presumably agree that one has such a right that
is (ordinarily) violated if one is prevented from voting.
Are any rights violated by the Kraft takeover of Cadbury? I do not think
that they are. Rodin describes Cadbury as being built along Quaker values
of dignity and respect, with a focus on providing a holistic approach to
employee welfare, including the provision of high-quality housing, subsidized
recreational activities, and funding service to the community. We can see
that the employees certainly had an interest in the continued provision of
these goods and the continued existence of this community. But not all of our
interests are protected by rights.
I’ve argued that, at the level of the state, successful political aggression
backed by conditional force effects a significant change in the sort of soci-
ety that citizens live in. It worsens their lives by depriving them of impor-
tant goods such as their freedom elect their government, to protest against
unjust treatment and so on. These goods are things to which people have
rights, and thus political aggression at the level of the state violates those
rights.
But this doesn’t seem true in the case of a takeover by Kraft. This is, in
part, because I don’t think the employees had rights to the continued provi-
sion of the goods provided by Cadbury. While we have a right to try to form
communities, rooted in our broader and very important right of free asso-
ciation, we do not obviously have rights to the continued existence of our
communities. If, for example, I start a War Ethics club, this might be an
enterprise that is very valuable to me, that benefits me in various ways and
so on. But if the other members of the War Ethics club foolishly decide that
they have better things to do with their time, I have no moral power at all
to try to coerce them into staying, even if their departure brings about the
collapse of my valuable community. What I have is, as I said, a right to free
association that includes a right to try to form communities and other rela-
tionships, but not a right to the continuation of any particular community
or relationship.
10. RIGHTS AND COMMUNITIES
It’s this right that is protected by the state, and that is not protected by a
group such as Cadbury nor threatened by a group such as Kraft. Notice that
it was perfectly possible for those ousted from Cadbury by Kraft to go and
set up another community, run along Quaker lines, elsewhere. Kraft might
be able to destroy the employees’ existing community, but it has no power
to prevent them from establishing other or replacement communities.
Importantly, it’s the state that provides the protected space in which those
employees can try to form a new association. I think this feature of a state
helps explain why states (and some other entities) can warrant lethal defense
while an entity like Cadbury cannot. The state protects not our particular
communities, but our right to form those communities.
We shouldn’t confuse this with the claim that the state is valuable because
it protects many communities, whereas a particular community like Cadbury
does not. Rodin considers and rejects this explanation of why states warrant
defense: “This proposal suffers from the problem [that if ] the state possesses
value because of the communities that exist within its territory, then we
would expect that states with rich ecosystems of local communities would
enjoy greater defensive rights than states with sparser communal activity.
They do not” (Rodin 2014: 74). Rodin is addressing, here, the idea that the
source of defensive rights lies in the value of the communities that are pro-
tected (hence, his claim that more communities ought to equal greater defensive
rights). But my claim is that the real value lies in the state’s protecting its
members’ rights to free association. Both the communities and the right are
valuable, but it’s the right that plays a role in establishing proportionate
defense.
This way of understanding the claim answers Rodin’s concern that states
with less communal activity will end up with less stringent defensive rights.
On this account, it’s not the communities per se that are defended, but the
right to form them. States that protect this right, along with its citizens’ other
important rights, are equally valuable in this respect, even if their members
choose to exercise their associative rights to different degrees.
There’s perhaps a simpler variant of Rodin’s reply that might be leveled at
my view—namely, that it implies that more populous states have more
stringent rights of defense than less populous states. After all, if the numbers
matter, then it seems that it will be proportionate for some states to do more
than others in the face of aggression, since they have a greater number of
members whose rights are threatened by aggression.
I think this is correct, but that this is not an objection to reductive individual-
ism. Rather, this seems to me an appealing account of defensive rights.
More populous states protect the interests of larger numbers of people, and
it’s therefore quite plausible to think that they may do more in defense of
those people. When the good protected is “larger” in this sense, it’s propor-
tionate to do more to protect it. There could be reasons why international
law ought not to dictate that some states may use only less harmful defense:
such a law would invite conditional aggression of a particularly violent sort
towards those states. But it may nonetheless be true as a matter of morality
that a state with a small population may not do as much to defend its mem-
bers’ political rights as a more populous state. But this is compatible with
thinking, as I have argued, that political aggression of the sort under discus-
192 Helen Frowe
sion here threatens serious rights, and that these rights are serious enough
that it will be proportionate for most (and probably all) states to use mili-
tary force to protect them.
11. CONCLUSION
The Conditional Force Argument raises an important challenge for the

reductive individualist. While reductivists have done extensive work on var-
ious aspects of jus in bello, our accounts of jus ad bellum remain under
developed. In this chapter, I have argued that we can explain why it’s
proportionate to resist political aggression even foreseeing that this will cause
an aggressor to wage a war that will expose many people to the risk of seri-
ous harm. These mediated harms, while not irrelevant to defenders’ propor-
tionality calculations, are nonetheless heavily discounted in those calculations.
Contra Rodin, this discount is not canceled out by a duty of care to one’s
fellow citizens. I have also suggested that our political rights are weightier
than proponents of the Conditional Force Argument have allowed.
Aggregated across a number of people, it can be proportionate to use force
to defend these rights. Moreover, once a sufficient number of people’s polit-
ical rights are threatened, the nature of the threat can change significantly.
We are no longer simply iterating the same harm across each member of a
group of individuals, but effecting a significant change in the sort of society
they live in.
However, if this kind of argument succeeds, it may still lead to the
Proliferation Problem—that is, to the claim that various non-state groups
are permitted to use force to defend their members against threats to the group’s
integrity. I argued that we can reply to this objection by noticing that merely
having an interest in the continuation of a community such as Cadbury
does not equate to a right to that community’s continued existence. Members
of states may act together to protect their right to form communities, but
it’s the protection of the right that is of central importance when it comes
to determining the proportionality of defense, rather than the protection of
the communities themselves. This answers the concern that states with
healthier communal lives may enjoy greater defensive rights. But it will not
provide us with a bright line between states and other entities when it comes
to rights of defense. It’s possible that other groups can act to defend not a
particular community’s integrity, but the right of its members to form com-
munities. The members of such groups could, therefore, also have be per-
mitted to use force to defend this right, even foreseeing that such force will
be met with further, more serious unjust threats.
Bibliography
Buchanan, Allen (1998). “Secession and Nationalism,” Goodin, R. and Pettit, P.
(eds), A Companion to Contemporary Political Philosophy (Blackwell).
Fabre, Cécile (2009). “Guns, Food and Liability to Attack in War,” Ethics, Vol. 120,
No. 1, 36–63.
Frowe, Helen (2014). Defensive Killing (Oxford: Oxford University Press).
Kamm, Frances M. (2014). “Self-Defence, Resistance and Suicide: The Taliban
Women,” in Helen Frowe and Gerald Lang (eds), How We Fight: Ethics in War
Kutz, Christopher (2005). “The Difference Uniforms Make: Collective Violence in
Criminal Law and War,” Philosophy and Public Affairs, Vol. 33, No. 2, 148–80.
Lazar, Seth (2014). “National-Defence, Self-Defence and the Problem of Political
Aggression,” in Cecile Fabre and Seth Lazar (eds), The Morality of Defensive War
McMahan, Jeff (2009). Killing in War (Oxford: Oxford University Press).
Rodin, David (2014). “The Myth of National Self-Defence,” Cécile Fabre and Seth
Lazar (eds), The Morality of National Defence (Oxford: Oxford University Press).
Tadros, Victor (2011). The Ends of Harm: The Moral Foundations of Criminal Law
Tadros, Victor (forthcoming), “Unjust Wars Worth Fighting For,” Journal of Practical
Ethics.
Walzer, Michael (1974). Just and Unjust Wars (New York: Basic Books).
Zohar, Noam (1993). “Collective War and Individualistic Ethics: Against the
Conscription of Self-Defense,” Political Theory, Vol. 21, No. 4, pp. 606–22.
8
Elbow Room for Rights
Eric Mack
1. INTRODUCTION
Robert Nozick’s first sentence in Anarchy, State and Utopia is the ringing
affirmation that “Individuals have rights, and there are things that no per-
son or group may do to them (without violating their rights).” Nozick imme-
diately adds that these rights may be even more restrictive than one might
anticipate. “So strong and far-reaching are these rights that they raise the
question of what, if anything the state and its officials may do. How much
room do individual rights leave for the state?” (1974, p. ix).
In this chapter, I address questions about whether core rights affirmed by
Nozick, viz., persons’ rights over themselves and over their legitimately acquired
extra-personal holdings, are considerably more deeply restrictive than Nozick
himself anticipates when he wonders whether these rights would morally hog-
tie aspiring statesmen and state officials. Indeed, I address questions about
whether those rights are so constraining that they deeply restrict their own
exercise. These questions have been forcefully pressed in two important critical
discussions of rights-oriented libertarian thought, Peter Railton’s “Locke,
Stock, and Peril: Natural Property Rights, Pollution, and Risk” (1985) and
David Sobel’s “Backing Away from Libertarian Self- Ownership” (2012).
Both of these papers maintain that the rights of self-ownership and of
private property in extra-personal holdings—which for convenience I will
refer to as the “libertarian rights” or the “Lockean rights”1—are much more
restrictive than their advocates generally recognize. The key thought is that
libertarian rights demand that all agents desist from even the most minor
unprovoked and non-consensual physical intrusions upon right-holders; in
1
Although I employ these terms to refer to rights of self-ownership and individual
entitlements to extra- personal objects, I do not mean to say that these are the only moral
rights libertarians should affirm or that self-ownership is the only natural moral right.
Elbow Room for Rights 195
addition, these rights at least seem to demand that all agents desist from
imposing on others even modest risks of such intrusions. Yet, compliance
with these demands requires that individuals forego all or almost all of the
ordinary exercises of rights by which individuals maintain their lives and
achieve their ends. As legal theorist Richard Epstein puts it, “If any smell,
noise, or discharge counted as a nuisance [and, thus, a rights infringement],
no one could barbecue in the backyard, talk on his front patio, or farm”
(2009, pp. 15–16).
The distinctive complaint that Railton and Sobel bring against stringent
libertarian rights is that such rights morally hog-tie their possessors. The
purported difficulty is that these rights systematically make their exercise
morally impermissible. Railton signs on to this complaint when he says,
I will argue that when one attempts to apply such [Lockean rights] theories to moral
questions about pollution, they present a different face, one set so firmly against
laissez faire – or laissez polluter – as to countenance serious restriction of what
Lockeans have traditionally taken to be the proper sphere of individual freedom.
(1985, p. 89)
Sobel signs on to this complaint when he says,
The ubiquity of difficult to avoid, minor infringements on other people’s bodies makes
the simple argument [from libertarian rights to standard libertarian conclusions]
unattractive. Strong moral constraints against all such infringements would make
too many things impermissible. The thought that, quite generally, my self-ownership
creates very powerful moral constraints on any and all involuntary infringements on
my body would unacceptably interfere with your liberty as Nozick saw. (2012, p. 35)
Could the philosophical theory named for liberty actually turn out to be unacceptably
restrictive of our liberty? (2012, p. 37)
The hog-tying complaint is not that the domains of some individuals may be
too small; i.e., that they may include too little in the way of personal or extra-
personal resources. Rather, the complaint is that much of whatever is within
one’s domain in the sense that others may not destroy or seize or control it
without one’s permission will, nevertheless, not be within one’s sovereign
domain in the sense that one’s chosen use of that material is morally allowed
and protected. Note that this complaint can be directed against any doctrine
that broadly ascribes rights to individuals over persons or extra-personal
objects that others are morally forbidden to infringe. For example, if rights
to extra-personal objects are ascribed to individuals on the basis of a strongly
egalitarian-leaning end-state principle and these rights are taken to be
infringed by any unconsented to physical intrusion, those rights will be subject
to the hog-tying complaint.
In this chapter, I focus on what the friend of libertarian rights or any other
advocate of similarly potent rights should say about “minor intrusions,” i.e.,
196 Eric Mack
impositions of very low-level physical effects upon another person or her

property. I shall bypass the more complex companion issue of what should
be said about the creation of pure risks of intrusions; i.e., the generation of
risks of intrusions that do not eventuate in actual intrusions2 although the
expected intrusiveness is comparable to the actual intrusiveness of minor
intrusions. I take it that any sensible friend of libertarian rights will want
to say that minor intrusions are morally permissible. Ceteris paribus it is
permissible for me to drive my car up my driveway even though this causes
a slight vibration in my neighbor’s eardrums to which he has not consented.
The difficulty is to provide an explanation for this permissibility that does
not render rights less morally imposing or less contra-consequentialist than
libertarian rights theorists want them to be.
Railton and Sobel maintain that a natural move for the friend of libertar-
ian rights involves a general shift to a more attenuated, less demanding,
understanding of rights. More specifically, this is a shift from an understand-
ing of claim-rights as moral claims that are protected by property rules to an
understanding of claim-rights as moral claims that are protected by liability
rules. If a moral boundary is protected by a property rule, others are simply
morally required not to cross that boundary without the consent of the
right-holder. On the other hand, if a boundary is protected by a liability rule,
others may cross that boundary without the consent of the right-holder as
long as due compensation is paid. A property rule says “do not cross this line”
while a liability rule says “crossing this line makes you liable for paying due
compensation.”3 If one is drawn to both B having a right that seems to exclude
A from sending smoke from his barbecue on to her property or into her eyes
and to A having the right to barbecue in his backyard, the shift to the liability
rule construal of rights may seem attractive. For this approach neatly splits
the difference. A gets to barbecue as long as he offers B a compensating half-
slab of ribs.
Indeed, Railton and especially Sobel take Nozick himself to shift to a liabil-
ity rule construal of rights in his chapter on “Prohibition, Compensation, and
Risk” (1974, pp. 54–87) at least in part to secure the permissibility of minor
intrusions (and the generation of moderate pure risk).4 Railton and Sobel
correctly see this liability rule attenuation of rights as an abandonment of
or “backing away” from the robust and strongly anti-consequentialist
2
On pure risks, see Railton, p. 95.
3
The now common “property rule” and “liability rule” terminology is introduced in
Calabresi and Melamed.
4
Sobel (p. 38) says that Nozick’s “main response [to the apparent impermissibility of
minor intrusions] is to claim that our property rights do not create boundaries that it is
generally impermissible to cross. Rather, others may permissibly cross our boundaries
provided that they adequately compensate us for doing so . . .”
understanding of rights that is expressed in that opening sentence of Anarchy,

State and Utopia. They welcome this attenuation of rights partially because
they see it as weakening the Lockean’s resistance against socially beneficial
infringements of rights that are not accompanied by compensation to the
subject of the infringement. So, e.g., Sobel says,
. . . seemingly a wide range of actions that involve taking from those who will little
feel the loss and giving to those seriously and nonculpably in need would be
permissible for the same reason some pollution and trivial risk is permissible – namely
because the infringement harms are trivial and the social benefits great. (2012, p. 60,
emphasis added)
I agree that a general liability rule attenuation of rights explains the permissibility
of minor intrusions in a way that undermines the robustness of the rights
that the libertarian theorist endorses. Hence, the need for the friend of
libertarian rights to offer an alternative explanation for the permissibility of
minor intrusions. Still, I do not think that liability rule attenuation explanation
readily leads to uncompensated boundary-crossings being permissible as long
as the “harms are trivial and the social benefits great.” Moreover, the liability
rule attenuation of rights does not introduce a “utilitarianism of rights”
(Nozick 1974, p. 28). A still may be required not to infringe upon B’s rights
without duly compensating B even if this infringement without compensation
is necessary to prevent the infringement without due compensation of the
comparable rights of C and D. The friend of libertarian rights need not resist
liability rule attenuation in order to avoid these further depreciations of the
stringency of rights.
I offer an alternative “elbow room for rights” explanation for the permissibil-
ity of minor intrusions. The key idea is that, when one thinks about how to
articulate or delineate the character or the boundaries of the rights one ascribes
to persons, one crucial guide is the moral elbow room postulate.5 According
to this postulate, a reasonable delineation of basic moral rights must be such
that the claim-rights that are ascribed to individuals do not systematically
preclude people from exercising the liberty-rights that the claim-rights are
supposed to protect. When Railton and Sobel point out that the impermissi-
bility of minor intrusions would be hog-tying, they are pointing out that
this impermissibility would systematically morally preclude individuals from
exercising the liberty-rights that are ascribed to them—the exercise of which
is supposed to be protected by the claim-rights ascribed to them. The elbow
room postulate tells us that, since the impermissibility of minor intrusions
would be hog-tying, a reasonable delineation of rights does not construe
minor intrusions as boundary-crossings. Since minor intrusions are not to
5
I have previously labeled this “the anti-paralysis postulate.” See Mack (2012), pp. 112–14.
198 Eric Mack
count as boundary-crossings, no liability rule attenuation of rights is needed

to render those actions permissible. The permissibility of minor intrusions
is explained on the basis of a refinement in the location of boundaries rather
than a general attenuation of rights.
One might, of course, affirm the permissibility of minor intrusions on the
basis of some sort of calculus of overall social costs and benefits or on the basis
of everyone on net being better off if minor intrusions are permissible—
perhaps through explicit compensation being paid to those intruded upon
or (as we shall see) through what Epstein calls “implicit in-kind compensa-
tion.” However, unless there is an explanation within rights theory for that per-
missibility, any such affirmation will mark the general subordination of
rights theory to or its displacement by such a social calculus or such an
appeal to mutual advantage. That is why it is important for me to show that
the proposed refinement in boundaries is a matter of working out the impli-
cations of the underlying rationale for ascribing moral rights to individuals.
Then the overall social benefits or the mutual benefits of the permissibility
of minor intrusions can be seen as the not merely accidental by-products of
their independently established permissibility.
Furthermore, I argue that the moral elbow room postulate is not an ad
hoc stipulation dreamed up solely to deal with the difficulties that Railton
and Sobel explore. Rather, the postulate is well anchored in the underlying
reasons for affirming Lockean rights. The rationale for the deployment of the
postulate in the delineation of people’s basic rights emerges from the ration-
ale for ascribing people rights in the first place. Moreover, the postulate (or
something close to it) has been put to work or can appropriately be put to
work by Lockean theorists to deal with a range of over-restrictiveness issues
that are not considered by Railton or Sobel.
As I view it, the most basic organizing principle for Lockean rights theory
is the moral principle that each individual is to be allowed to live his own
life in his own chosen way.
Lockean rights theory needs both to provide philosophical support for
this organizing principle and to articulate it in terms of the ascription to
persons of an array of abstract moral rights. Each abstract moral right pro-
vides individuals with moral protection against one of the diverse ways in
which they can be prevented (by others) from living their own lives in their
own chosen ways (Mack 2010). For example, people can be prevented from
living their own lives in their own chosen ways by being deprived of dis-
cretionary control over their own bodies and faculties. For this reason, the
abstract right of self-ownership is a crucial and salient dimension of the
proper codification of the primordial libertarian principle. Also, people can
be prevented from living their own lives in their own chosen way by being
deprived of the opportunity to acquire and exercise discretionary control
over extra-personal objects. For this reason, the abstract right of property—a
right to make things one’s own and to exercise discretional control over
what one has made one’s own—is also a crucial and salient dimension of
the proper articulation of the libertarian organizing principle.6
Is an individual who needs a kidney disallowed from living her own life
in her own chosen way by another individual declining to supply her with
his spare kidney or by his evading her attempts to extract that needed kidney?7
Is an individual who desires to engage in certain sexual acts precluded from
living his own life in his own chosen way by the sole object of his sexual
desires declining to participate in those acts or by her evasion of his advances?
The organizing principle for Lockean rights answers these questions in the
negative. Agents who are not enabled to live as they choose because another
party declines to supply one of his kidneys or declines to participate in
desired sexual interactions are not thereby made unable—or precluded from—
living their own lives in their own chosen ways. “Failing to help another
cannot be construed as interfering with his right to use himself as he
wishes . . .” (Cohen 1995, p. 215). Absent this understanding, ordinary exer-
cises of rights by one party will regularly also count as ordinary violations of
other parties’ rights and the moral claims of individuals to live their own
lives in their own chosen ways will systematically conflict. Any resolution of
that conflict would require the demotion of rights into moral commodities
that are to be traded off against one another.
I need to emphasize that it is not the role of armchair philosophy—even
natural rights philosophy—to discover and disclose the precise contours of
persons’ nitty-gritty rights. Those precise contours do not exist out there in
the nature of things or as theorems that are deducible from Lockean axi-
oms. So, it is not the business of a Lockean theory of rights to determine
whether or not the owner/operator of a well-established water mill has a
right against individuals living upstream that they not significantly dimin-
ish the flow of water that turns his mill. It is not the business of this or any
other philosophical theory to determine exactly how loud the noise has to
be that emanates from A’s property in order for B to have a right to enjoin
A’s drum-playing.
The relatively concrete rights that are reasonably ascribed to individuals
in a given society—e.g., the right not to be subjected to noise over a certain
decibel level—provide a structure of reasonably expected liberties and
immunities that facilitate peaceful coexistence and voluntary cooperative
6
A right against deceptive manipulation would be a third element within the abstract
framework of libertarian rights. For deceptive manipulation is yet another way in which
one can be precluded from living one’s own life in one’s own chosen way.
7
I thank Sobel for pressing this question.
200 Eric Mack
interaction among those individuals. A theory of rights can only provide an

abstract framework for that relatively concrete structure.
There are two phases in the articulation of that abstract framework. The
first of these is the articulation of abstract rights, each of which constitutes
a moral barrier against one of the basic ways in which persons’ underlying
moral claim to be allowed to live his own life in his own chosen way can be
infringed. The second phase involves further investigation about where the
boundaries that define those rights may lie and what sort of stringency
those rights must have or must usually have. This chapter operates within
that second phase by focusing on whether or not minor intrusions are to be
construed as boundary-crossings within a reasonable delineation of Lockean
rights. Beyond the philosophical articulation of that abstract framework is
the actual concrete instantiation of abstract rights in people’s judgments
and in the rules and practices that people are disposed to follow at least in
part because they expect others to follow them. While the articulation of
abstract Lockean rights rules out many particular judgments, rules, and
practices, it does not single out one set of particular judgments, rules, and
practices—one particular set of concrete rights—as that which is required
by these abstract rights.8 This is why it is a mistake to expect Lockean rights
theory to disclose precisely what procedures are needed to acquire or trans-
fer a property right or to spin out the details of justifiable riparian or nui-
sance law. From the point of view of Lockean theory it is enough for existing
judgments, rules, and practices to be within the range of the acceptable and
for them to guide conduct in accordance with the function of rights.9
I proceed as follows: in section 2, I focus on how unattractive general
liability rule attenuation is for the Lockean theorist and, hence, how eager that
theorist should be to find a different explanation of the permissibility of
minor intrusions. In section 3, I explore the explanation for the permissibil-
ity of minor intrusions advocated by the Locke-leaning legal theorist
Richard Epstein.10 Epstein’s scheme allows minor intrusions on individuals
on the basis of those individuals being implicitly compensated in-kind by
being allowed themselves to engage in minor intrusions on others. Epstein’s
scheme is of interest for three reasons. First, if the Lockean cannot persua-
sively argue that minor intrusions are permissible because they are not
boundary-crossings, Epstein’s version of liability rule attenuation seems like the
best fall-back position. Second, contrasting Epstein’s in-kind compensation
8
This sort of layered articulation of rights is suggested by Loren Lomasky’s distinction
between basic rights, moral rights, and legal rights (Lomasky 1987) and Gerald Gaus’
distinction between abstract moral principles and more concrete moral rules (Gaus 2012).
9
See Mack (2010), p. 63 and Gaus (2012), pp. 391–417.
10
See Epstein (1979) and (2009).
proposal with my elbow room proposal serves to clarify the character of

the latter stance. Third, developing this contrast also points to the greater
plausibility of the elbow room proposal.
Epstein draws our attention to Baron Bramwell’s invocation in Bamford
v. Turnley (Bamford 1862) of a live-and-let-live maxim in support of the
permissibility of minor intrusions. I examine Bramwell’s stance in sec-
tion 4 and contend that it is at least as readily interpreted as an elbow room
argument as an implicit in-kind compensation argument. Bramwell also
maintains that minor intrusions—what he calls “annoyances”—that are
wanton or malicious may be enjoined even though those minor intrusions
are to be allowed if they are performed non-wantonly and non-maliciously.
In section 5, I sketch how implicit in-kind compensation reasoning and
elbow room reasoning can each underwrite this distinction. This distinc-
tion enables the Lockean rights advocate to hold—as Nozick held (1974,
p. 75)—that stealing a penny is impermissible without also holding that
incidentally causing a penny’s worth of damage to another’s property is a
boundary-crossing. In section 6, I expand on my claims that elbow room
delineation of rights is well rooted within Lockean natural rights theory
and I explain why elbow room reasoning is not, contrary to appearances, a
species of implicit in-kind compensation reasoning. In section 7, I con-
clude with an observation or two about the relationship between armchair
philosophical delineation of abstract rights and the assessment of actual
concrete legal rights.
One final note about my use of “minor intrusions” may be helpful. Some
intrusions may be so minor that their permissibility does not really require
any refinement of the rights of those subject to those intrusions. These are
the intrusions that have no discernible effect on the discretionary control
that persons have over the objects of their respective rights. So, if persons C
and E communicate with one another by means of certain radio waves that
pass through D’s home or person yet those waves have no physical effect on
D’s home or person, we may say either that their communication activities
are not “intrusive” or that they are “intrusive” in a sense that does not raise
any question about their permissibility. Recognizing this may ease the way
to accepting that minor intrusions that are not quite so de minimis—e.g.,
causing a few smoke particles to land on one’s neighbor’s lawn—may also
not be rights infringements.
2. ASSESSING LIABILITY RULE ATTENUATION
For the advocate of libertarian rights, the construal of persons’ most

fundamental moral rights as claims protected by liability rules fails to
202 Eric Mack
capture much of what makes those rights attractive. For central to the
appeal of rights is that they provide individuals with moral protection for
their choices with respect to a range of alternatives. If one has rights over
oneself and over one’s holdings, one may dispose of oneself or one’s holdings
as one sees fit; one may do as one wills—subject, of course, to the standard
requirement that one is not violating others’ rights. As Locke held, each
person’s fundamental claim against others is to “. . . a Liberty to dispose, and
order as he lists, his Person, Actions, Possessions, and his whole Property,
within the Allowance of those Laws under which he is, and therein not to
be subject to the arbitrary Will of another, but freely follow his own” (1689,
Second Treatise, §57).
All the standard vocabulary of individual sovereignty, moral inviolability,
moral space, individual spheres of authority, and domains of personal free-
dom reflect the really basic idea that rights are about the protection of
choice. Persons may be harmed in certain ways when their rights are vio-
lated and that harm may be counterbalanced by simultaneous or subse-
quent due compensation; but such compensation does not counterbalance
the wrong that is done to the right-holder by depriving him of the choice of
whether he will be subject to that counterbalanced harm. The slate is not
wiped clean by due compensation because compensation deals with the
harm and not the wrong.
What if there is some extra compensation for being deprived of the choice
about whether one will be subjected to the counterbalanced harm? This
question suggests that having choice in the matter is itself just another inter-
est of the right-holder that is not to be set back without due compensation.
But even if we say that having choice in the matter is an interest of the right-
holder, it remains a very special sort of interest. For the claim to having the
choice about whether to be subjected to counterbalanced harm cannot be
honored by depriving the claimant of that choice and throwing in a bit of
extra payment. David Schmidtz rightly emphasizes that at the core of a
property right—indeed, at the core of any moral claim protected by a prop-
erty rule—is “a right to say no to proposed terms of exchange” (2010, p. 79).
One’s right to say no is not honored by others infringing upon that right
while providing one with the payment one would have accepted if one had
not said no. One’s right to say no is not honored if the treatment one objects
to is imposed upon one—even if it is true that, had one deliberated well,
one would have waived one’s right against the treatment in exchange for the
payment that is now offered as compensation.
A general liability rule attenuation of rights does not comport with
Nozickian invocations of moral inviolability (1974, pp. 31–2). One salient
dimension of belief in the moral inviolability of persons is subscription to
principled anti-paternalism. According to principled anti-paternalism, at
least some interventions that aim at the recipient’s good and on balance
do advance the recipient’s good are still impermissible infringements of
the recipient’s rights. However, liability rule attenuation undermines prin-
cipled anti-paternalism. For, given that attenuation, if an intervention
bestows an over-balancing good on its recipient, there will be no infringe-
ment of rights.
Nozick seeks to elucidate moral inviolability by contrasting beings who
possess inviolability with tools (1974, p. 31). But, how much less of a tool is
one if one has rights that are protected by liability rules rather than having
no rights? Suppose that there were a moral norm according to which anyone
who uses a particular unowned hammer is morally required to repair any
damage done to it or even to shine the hammer up a bit. It does not seem
that such a norm would make the hammer any less of a tool than it would
be without protection by that liability norm. What more is present when
everyone else is required to duly compensate you—to repair any damage to
you or even shine you up a bit—after making use of you? The only thing
more that is present is that you—unlike the hammer—have a claim to that
compensation. Not only is there a price for using you; the payment of that
price is owed to you. So, your having a right against being used in certain
ways that is protected by a liability rule is a little bit more than there being
a moral norm that says that whoever makes use of you in an intrusive way
must repair you or even shine you up a bit. Nevertheless, it is very difficult
to see that this little bit more makes you significantly less of a tool than that
unowned hammer.
In Anarchy, State and Utopia, Nozick devotes several trenchant pages to
his rejection of H. L. A. Hart’s principle of fairness (1974, pp. 90–5).
According to this principle, individuals who have taken on certain bur-
dens in the course of some benefit-conferring activity have a right against
all beneficiaries of that activity that they share in those burdens. So, if you
have benefited from all your neighbors taking time to tell funny stories on
the neighborhood public address system, you are bound yourself to spend
some time telling funny stories (or the like). Nozick rejects this saying,
“One cannot whatever one’s purposes, just act so as to give people benefits
and then demand (or seize) payments” (1974, p. 95). Yet, if one accepts
the liability rule attenuation of rights, why shouldn’t this be allowed?
Why isn’t the benefit already conferred by those who now demand (or
seize) payment simply advance compensation for the boundary-crossing
they subsequently engage in? If one may seize as long as one makes due
compensation afterwards, why may one not make due compensation for
a future seizure?
As one would expect, a general liability rule attenuation of rights renders
many actions morally permissible that the friend of libertarian rights plausibly
204 Eric Mack
takes to be violations of rights.11 Ceteris paribus all forced exchanges that leave
their subjects at least as well off as they would be were they not subject to
the forced exchange are rendered permissible—whatever the absolute mag-
nitude of the imposed costs and compensating benefits. Consider just two
fairly extreme examples. The first is forced participation in eyeball (or kid-
ney) redistribution pools. All members of the pool start with two healthy
eyeballs (or kidneys). If a member loses both eyeballs (or both kidneys),
some other member of the pool is required to donate an eyeball (or kidney)
for that first member. The details for membership in a given pool are set so
that, for each individual who is required to enter that pool, the expected loss
from being a member is comfortably less than the expected gain. So, the
pool is one that rational individual cost-benefit calculators would freely
enter. But imagine further that many people simply do not voluntarily sign
up for their designated pools. Liability rule attenuation seems to allow forc-
ing people into such pools on the simple grounds that the (expected) cost of
participation is more than duly compensated for by the (expected) gain.
The second example takes advantage of the fact that liability rule compen-
sation need not be direct or intended. Imagine that a number of individuals
are captured by slave raiders, transported to another continent, and held in
pretty bad slavery for the rest of their lives; but, had they not been captured,
they would have almost immediately been wiped out by an unanticipated
and horrendous disease. Even though we may blame the slave raiders for
crossing boundaries without anticipating the accompanying due compensa-
tion, under liability rule attenuation no rights of the enslaved turn out to be
violated. The enslaved would have no claims to any (further) compensation
from the raiders; and, of course, there would be no rights-violation basis for
punishing the raiders.12
3. EPSTEIN ON IMPLICIT IN-KIND COMPENSATION
In this section, I present Richard Epstein’s implicit in-kind compensation

account of the permissibility of minor intrusions. According to Epstein
(2009, p. 16), “To head off those results [of barbecuing, talking, or farming
11
A thorough discussion of Nozick’s view would examine the ways in which, according
to Nozick, difficulties in determining subsequent due compensation for boundary-crossings
may lead to the conclusion that many boundaries ought to be treated as though they are
protected by property rules. For a condensed discussion, see Mack (2012), pp. 100–3.
12
Railton (1985, p. 115) and Sobel (2012, pp. 46–7) also criticize liability rule
attenuation for not sanctifying the choice of the right-holder even though they themselves
would allow overriding the choices of right-holders without compensation when this is
sufficiently socially beneficial.
being prohibited], a strong live-and-let live principle allows all low-level

nuisances to continue without [explicit] compensation, and this creates
universal Pareto improvements that should be welcomed on all sides.”
I examine Epstein’s stance because it seems to be the best case for the
permissibility of minor intrusions that is based upon (something like) a
liability rule attenuation of rights and because consideration of Epstein’s
stance allows me to develop the contrast between this implicit in-kind
compensation view and the elbow room rationale for the permissibility of
minor intrusions.
We should notice, however, that Epstein’s stance does not quite amount to
a liability rule attenuation of rights. Epstein holds that there are very substan-
tial “utilitarian” gains associated with allowing minor intrusions; yet, he holds
these gains do not themselves vindicate (even) minor intrusions. For these
intrusions are affronts to personal autonomy; and our rights—including our
private property rights—are supposed to protect autonomy.
. . . one of the essential functions of private property [is] that of specifying for each
person a domain of action in which he is not accountable to the whims or the
demands of any other group. Property is an external manifestation of the principle
of personal autonomy. (1979, p. 63)
Nevertheless, Epstein concludes that the wrongfulness of subjecting an
individual to such intrusions is sufficiently tempered by that individual’s
enjoyment of in-kind compensation; i.e., compensation in the form of that
individual’s own opportunity to engage in minor intrusions.
[C]ompensation from beneficiary to victim should in general be required, because
forced exchanges represent less of an affront to corrective justice principles than do
outright takings.” (1979, p. 77, emphasis added)
This forced rearrangement of rights does compromise individual autonomy, because
it abridges the right of all to decide whether to retain or to dispose of what they own.
But its improper effect is softened because it [i.e., the general policy of allowing non-
consensual crossings] is aimed at all for the benefit of all, and not at A alone for the
sole benefit of others. (1979, p. 78, emphasis added)
Minor intrusions are to be allowed primarily because the general allowance of
such intrusions provides each individual who is subject of these intrusions with
in-kind compensation; they are to be allowed even though these compensated
for intrusions seem to remain somewhat morally tainted. Presumably more
extensive intrusions are also to be allowed if those subject to them receive the
due in-kind or out-of-kind compensation as that would be required under a
general liability rule attenuation of rights.
Whereas Railton points to difficulties in a policy of case-by-case monetary
compensation of individuals for particular minor intrusions (1985, p. 108),
Epstein’s distinctive claim is that no such program of monetary compensation
206 Eric Mack
is needed because (at least typically) each individual gains more in welfare
from being allowed to engage in minor intrusions than she loses in welfare by
being subject to them.13
A is worse off to the extent that he cannot do with his land what he could have done
before [minor intrusions were allowed] . . . Yet by the same token he is better off to
the extent that the same regulation binds all others (B through Z) [to allow A’s
intrusions] for his benefit. The parallel restrictions upon others become implicit
in-kind compensation for A, and likewise for all others in the group. With
compensation thus assured there is no need to undergo the expensive and pointless
process of making explicit offsetting payments, whose sole effect is to dissipate the
welfare gains generated by the change in legal rules [from those that forbid minor
intrusions]. (1979, p. 78)
Epstein’s implicit in-kind compensation position is one version of a common
response to the restrictiveness of prohibitions on minor intrusions, viz., that
surely we are all willing to be subject to a whole host of minor intrusions in
order ourselves not be required to avoid engaging in such intrusive activities.
However, it is crucial to distinguish between Epstein’s in-kind compensation
version and the elbow room version of this response. According to the
implicit in-kind compensation version, what makes minor intrusions upon
B permissible is that the actions that will be performed by B and by others if
minor intrusions are permissible will render B better off than (or at least as
well off as) B would be rendered by the actions that would be performed by
B and others if minor intrusions are forbidden. What matters is the actual
outcome for B’s utility or welfare of the introduction of the permissibility of
minor intrusions. Thus, Epstein declares that “. . . the acid test is whether the
overall scheme provides implicit in-kind compensation by allowing each
landowner an appropriate fraction of the resulting gain” (2009, p. 28, emphasis
added). In contrast, according to the moral elbow room version, what makes
minor intrusions upon B permissible is that only if such intrusions are
permissible will individuals (including B) systematically have discretionary
control over the domains that are ascribed to them as a matter of right. What
matters is that people’s rights be delineated in a way that systematically allows
individuals to dispose of their persons and property as they respectively see
fit. Thus, the permissibility of A’s engaging in minor intrusions upon B is
not hostage to B actually employing his moral liberty to engage in minor
13
Epstein could further fortify this position by pointing out that (typically) individuals
gain enormously from others being allowed to impose minor intrusions on one another.
A gains knowledge about these things called fireplaces that can keep one from freezing to
death from B is allowed to build and use a fireplace even though that deposits some ash
onto C ’s property. A is enabled to trade with B because B is allowed to drive her wagon
past C even though doing so vibrates C ’s eardrums. And so on.
intrusions in ways that render B better off than (or at least as well off as) B
would be were minor intrusions forbidden.
Indeed, Epstein’s own statement of his position straddles the fence between
the in-kind (utility or welfare) compensation and the elbow room versions.
In the first three sentences of the passage cited immediately above, Epstein
seems to say that, although A has to allow minor intrusions on his land (e.g.,
the depositing of some smoke particles), others in turn have to allow his
minor intrusions upon them. The general restriction against prohibiting
minor intrusions leaves each better off with respect to their liberty to do as
they see fit with their property. While a recalibration of rights that construes
minor intrusions to be permissible involves a minor contraction of the
boundaries of A’s rightful domain, the associated permissibility of A’s minor
intrusions secures A’s discretionary sovereignty over that slightly retrenched
domain. Only in the final sentence in this passage does Epstein clearly express
the implicit in-kind (utility or welfare) compensation stance. Only in that
sentence is he clearly thinking about how individuals will actually act given
the recalibration of rights and is he vindicating the recalibration on the grounds
that the anticipated actions will on net advance the utility or welfare of each
of the parties.
We can see a similar fence-straddling when Epstein argues that the rights
of landowner A should not be understood as rendering it impermissible for
neighbor B, to make changes on her property that will diminish A’s view.
Beyond arguing that B’s disturbing A’s vista does not even satisfy the
“Physical Invasion Test” (1979, p. 60) for being a boundary-crossing,
Epstein offers the following consideration.
A’s claim to an unobstructed view is attractive only because it is considered in vacuo.
Yet the uniform protection of all views commits us to a set of entitlements that make
it impossible for anyone to use the land from which he might choose to look. (1979,
p. 61)
Although Epstein certainly thinks of this as an implicit in-kind compensation
argument, its focus is not on each party actually gaining (utility or welfare)
from the actions that will ensue if vista alteration is allowed. Rather, the
focal point seems to be the elbow room consideration that, only if vista
alteration is allowed, will individuals have the discretionary control over
their persons and property that we expect them to have when their
ownership rights are affirmed.
Suppose that, even if vista alteration is allowed, landowner B will never
do anything on her land that will diminish neighbor A’s existing vista; but
A will engage in activities on his land that will diminish B’s view. Under
these circumstances, implicit in-kind compensation reasoning does not
yield the permissibility of vista alteration while elbow room reason does.
208 Eric Mack
4. BRAMWELL AND THE LIVE-AND-LET-LIVE RULE
In support of his implicit in-kind compensation stance, Epstein cites Baron

Bramwell’s famous opinion in Bamford v. Turnley. The operation of Turnley’s
brick kiln introduced noxious fumes into neighbor Bamford’s home.
Bamford appealed a lower court’s finding that Turnley need not compensate
Bamford for the damaging fumes. In the Court of Exchequer Chamber, the
lower court’s decision was overturned by a vote of five to one. Turnley was
found liable to pay compensation to Bamford. The majority relied upon the
general doctrine that, “No man may use his right so as to damage another;
though, on the other hand, every one has a right reasonably to use his
property, even if he should thereby annoy his neighbor” (Bamford 1862, p.
29). The majority believed that, on a proper understanding of people’s
rights, they could hold Turnley liable for the damage he had done while still
upholding the permissibility of annoying (i.e., engaging in minor intrusions
upon) one’s neighbors. However, the one dissenting judge, Pollock, argued
that, if Turnley was liable for the harm done by his noxious fumes, “anything
which, under any circumstances, lessens the comfort or endangers the health
or safety of a neighbour, must necessarily be an actionable nuisance”
(Bamford 1862, p. 31). In effect, Pollock argued that the majority had no
basis for drawing a line between damaging and annoying conduct; if they
found Turnley liable for his damaging action, they were committed to the
hog-tying conclusion that all annoyances were actionable nuisances.
Bramwell’s separate opinion in support of the majority was largely devoted
to meeting Pollock’s challenge. Bramwell maintained that there is a class of
actions like “burning leaves, emptying cess-pools, and making noises during
repairs” that “may be lawfully done” even though they “would be nuisances
if done wantonly or maliciously”14 (Bamford 1862, p. 32). He recognized
that, in order to draw a line between non-wanton/non-malicious burning of
leaves, emptying cess-pools, and making noises during repairs and damag-
ing actions, he had to identify “some principle on which [the former] cases
must be excepted.” Bramwell then argued,
It seems to me that that principle may be deduced from the character of these cases,
and is this, viz., that those acts necessary for the common and ordinary use and
occupation of land and houses may be done, if conveniently done, without
submitting those who do them to an action . . . It is as much for the advantage of one
owner as of another; for the very nuisance the one complains of, as the result of
the ordinary use of his neighbor’s land, he himself will create in the ordinary use
of his own, and the reciprocal nuisances are of a comparatively trifling character.
We will consider shortly why wanton or malicious annoyances may be disallowed.

14
The convenience of such a rule may be indicated by calling it a rule of give and take,
live and let live . . . (Bamford 1862, pp. 32–3)
Epstein takes this passage to be an enunciation of the implicit in-kind
compensation stance. I think it is more ambiguous than that.
The majority of the justices—with whom Bramwell concurred—went out
of their way to say that A’s common and ordinary use and occupation of his
land and houses is “conveniently done” when that use or occupation is no
more than “an annoyance” to other parties (Bamford 1862, p. 30). Given this
understanding of “conveniently done,” the first sentence in the passage from
Bramwell seems to say that rights to land and houses must include rights to
dispose of those holdings even if one’s chosen action makes the residence
of another “less delectable or agreeable” (Bamford 1862, p. 30). We seem to
have a claim about the moral elbow room that must accompany one’s rights
to one’s lands and houses. If one has such rights, one must be at liberty to
use those holdings as one sees fit even if one’s exercise of that liberty makes
others’ holdings less delectable or agreeable. In contrast, the latter part of
the passage seems to compare the utility or welfare gains and losses for
individuals from their actually engaging in (non-malicious, non-wanton)
annoying conduct and being subject to such conduct. Since the losses from
being subject to the annoying conduct will be trifling and the gains from
engaging in the conduct will be more than trifling, a rule that allows each
to engage in annoying conduct is advantageous to all.
Bramwell provides a “live and let live” and a “give and take” synopsis of his
stance. The live-and-let-live synopsis fits nicely with the moral elbow room
reading. Each party has a right to live and more concretely to live by way of
discretionary control over his own person and possessions. Yet, each person’s
right of discretionary control over her person and possessions would come
to (almost) nothing if each person’s right included a right against imposed
annoyances. If each of us is to live, we each have to let everyone else live; and
this requires everyone having the moral liberty to annoy; i.e., the moral
liberty to engage in minor intrusions. This reading of the “live and let live”
synopsis fits especially very nicely with the general principle that the majority
relied in Bamford v. Turnley, viz., “No man may use his right so as to damage
another; though, on the other hand, every one has a right reasonably to use
his property, even if he should thereby annoy his neighbor”15 (Bamford
1862, p. 29). In contrast, the “give and take” synopsis fits better with
15
None of the judges who found for Bamford would allow Turnley to proceed with
his damaging fumes if he compensates Bamford. Bramwell says that a party who is
damaged has a claim to an injunction against the damaging activity. According to the
other judges who found against Turnley, if A’s operation of a tanning facility is damaging
to B, “it may be pulled down” (Bamford, p. 30).
210 Eric Mack
Epstein’s implicit in-kind compensation reading of Bramwell. Each gives

annoyance but, in turn, has to take it. Since it is better to give than it is bad
to receive, everyone in fact ends up better off.
What matters for the elbow room argument is that one attributes to people
the moral liberties that must be attributed to them in order for them to possess
the discretionary control over their lives and holdings that Lockean rights is
supposed to provide. This attribution is part of the reasonable articulation of
persons’ Lockean rights given the near impossibility of exercising rights over
oneself or one’s possessions without physically affecting to some extent other
persons or their holdings. What matters for the in-kind compensation
argument is that, if people are at liberty to engage in minor intrusions, they
will exercise that liberty in ways that will render everyone on net better off. It
is crucial for the compensation argument that the actual outcome of the
minor intrusions that will be performed will be Pareto superior to the out-
come that would obtain were actions forbidden. Due to these differences we
can imagine circumstances in which people are morally at liberty to engage
in minor intrusions according to the elbow room argument but not accord-
ing to the compensation argument.
Over-conscientious B would abstain from all actions on her property that
would convey noises, odors, or so on onto neighbor A or his property even
where minor intrusions are permissible. In contrast, A would engage in con-
venient uses of his land that would involve minor intrusions upon B. So the
actual outcome of minor intrusions being permissible within the mini-society
of A and B would be a net gain for A and a net loss for B. Hence, implicit
in-kind compensation reasoning would not allow minor intrusions in that
mini-society. However, elbow room reasoning would allow minor intrusions
on the now familiar ground that to construe the rights of person and property
as forbidding such intrusions would be to deny people the range of protected
liberty that their rights are supposed to secure.
In a somewhat different sort of case the net loss for B derives primarily
from her own chosen action. Suppose that B will burn herself alive if and
only if the minor intrusion that would consist of some of her ashes landing
on A or on A’s land were permissible. Our still very conscientious B would
forego that self-harming self-immolation if and only if sending those few
ashes onto A or onto A’s property were to count as a boundary-crossing. If
minor intrusions are permissible within the mini-society of A and B, even if
A abstains from minor intrusions upon B, B will be a net loser of utility or
welfare.16 So, under these circumstances also, the in-kind compensation
16
I assume that burning herself alive would be worse for B than starving—as B might
if B were too morally hog-tied to fertilize her fields or raise smelly pigs.
argument for A and B being at liberty to engage in minor intrusions will not
go through. In contrast, the moral elbow room argument for this liberty, viz.,
self-ownership would be rendered nugatory if persons were forbidden to dis-
pose of themselves in ways that were annoying to others, will still go through.
It seems to me pretty clear that it is permissible for A to motor up his
driveway even if some noise carries over to neighbor B who never herself
in fact engages in minor intrusions upon A. Similarly, it seems to me
pretty clear that B may permissibly cause some of her ashes to land on A
or his property in the course of her self-immolation even though it is
contrary to her interest for this pollution to be permissible. Also, it is
permissible for A to motor up his driveway even if B would self-immolate
were minor intrusions permissible. If one shares these judgments, one
should prefer the elbow room account of the permissibility of minor intru-
sions over the implicit in-kind compensation argument. The elbow room
argument better accords with “. . . a picture of a free society as one embod-
ying a presumption in favor of liberty, under which people permissibly
could perform actions as long as they didn’t harm others in specified ways”
(Nozick 1974, p. 78).
5. WANTON AND MALICIOUS INTRUSIONS
Bramwell maintains that the proper delineation of mine and thine leaves
room for actions that convey low-level odors or noises or vibrations to non-
consenting others; nevertheless, the conveyance of the same odors or noises
or vibrations will count as unlawful nuisances if “wantonly or maliciously”
done (Bamford 1862, p. 32). This, I think, is probably the stance that the
friend of Lockean rights needs to take. For it allows the Lockean to hold
that “stealing a penny or a pin or anything from someone violates his rights”
(Nozick 1974, p. 75, emphasis added) without having to hold that an action
that incidentally does a penny’s worth of damage to someone’s property (or
incidentally creates a 0.1 probability of ten penny’s worth of damage) is a
rights violation. However, as Railton argues, it seems that the wanton or
malicious performance of an act of a given type will count as a boundary-
crossing only if the non-wanton and non-malicious performance of that act
counts as a (presumably less blameworthy) boundary-crossing (1985,
p. 102). The boundary-crossing character of my maliciously depositing
some ash on your backyard seems to depend upon the boundary-crossing
character of my non-malicious (and non-wanton) production of that
outcome. Thus, if the malicious minor intrusion is a boundary-crossing, so
too must be the non-malicious (and non-wanton) intrusion.
212 Eric Mack
Nevertheless, we can readily envision both implicit in-kind compensation

and elbow room reasoning for affirming the permissibility of non-malicious
minor intrusions while denying the permissibility of malicious minor intru-
sions. In each case, we get an explanation for why, when rights are more
finely calibrated, malicious minor intrusions count as boundary-crossings
even though non-malicious intrusions do not. The in-kind compensation
reasoning is that adding the permissibility of malicious minor intrusions to
a structure that already extends to all a liberty to engage in non-malicious
minor intrusions will not be on net advantageous to all. For (the reasoning
goes) at least some people will be net losers if malicious minor intrusions are
generally allowed—viz., those for whom there is little value in maliciously
intruding upon others or high disvalue in being subject to malicious intrusion
by others. The moral elbow room reasoning is that, while individuals must
be at liberty to engage in non-malicious (and non-wanton) minor intru-
sions if they are to be at liberty to dispose of their own persons and posses-
sions as they see fit, this liberty need not extend to malicious (or wanton)
minor intrusions. It suffices to solve the hog-tying problem that non-mali-
cious and non-wanton minor intrusions be permissible. As long as the
minor intrusions on others are incidental to the agent’s decisions about how
to deploy his person or property we reasonably view these deployments as
fundamentally exercises of that agent’s rights. However, if those intrusions
are wanton or malicious—done for or verging on being done for their
intrusiveness—they are more reasonably seen as the agent doing as he sees
fit with others or their property and, hence, as boundary-crossing.
6. ELBOW ROOM REASONING
Given our vulnerability to interferences by others, in order for people to

possess the moral liberty to do as they see fit with their persons and
property, they must have rights-claims over themselves and their property
that require others to leave them in discretionary control of themselves
and their rightful holdings. Persons’ moral liberty to do as they see fit
with their persons and property would come to nothing (or almost
nothing) were that liberty not protectively clothed by such claim-rights.
When Locke insists that there are natural moral constraints on how one
individual may treat another—not just Hobbesian liberties to act as one
desires—his argument is that only if such constraints exist will each
individual genuinely enjoy “a Liberty to dispose, and order as he lists, his
Person, Actions, Possessions, and his whole Property . . .” “For who could
be free, when every other Man’s humour might domineer over him?”
(1689, Second Treatise, §57).
At the same time, the lesson to be drawn from the hog-tying challenge is
that, in order for people to possess the moral liberty to do as they see fit with
their own persons and property, they must not be morally precluded from
engaging in minor intrusions that are integral to their discretionary control
over and disposition of their own persons and property. The generation of
some negative spillover effect cannot be a basis for shutting down persons’
chosen use of what is their own. If rights were delineated in such a way that
such negative spillover effects provide a moral basis for shutting down per-
sons’ chosen actions, rights would not serve their (deontic) purpose of
defining spheres of freedom within which individuals may act as they see fit.
If each person’s protected sphere were confined to conduct “which affects
only himself ” (Mill 1859, p. 11), that sphere would be vanishingly small.
The rights that we ascribe to individuals to provide them with moral protec-
tion for the exercise of their liberties must leave persons morally vulnerable
to minor intrusions if those rights are not to shrink those spheres into (at
most) pinpoints of protected freedom.
The elbow room postulate or something close to it plays a systematic role
within rights theory by guiding the reasonable delineation of rights. Here
I will cite two examples of such guidance. The first example is Locke’s argu-
ment against there being an original joint-ownership of nature. Locke
accepts Robert Filmer’s claim that, if the earth were originally the joint
property of all of mankind, the establishment of any private property—
indeed, any permissible individual use of any portion of the earth—would
require a compact among all the joint-owners that has never taken place and
will never take place (Filmer 1652, p. 234). Locke, however, takes this to be
a premise in an elbow room argument against the original joint-ownership
of the earth. For, if the earth were the joint property of all and Filmer’s claim
is correct, there would be no elbow room for individuals to exercise their
most fundamental natural right, viz., “the right everyone had to take care of,
and provide for their Subsistence” (Locke 1689, First Treatise, §87). For
(almost) any exercise of this fundamental right requires that individuals be
at liberty to acquire private property or, at the very least, to make use of
portions of the earth.17
In addition, according to Locke, private property rights themselves must
not on net diminish the opportunity of individuals “to take care of, and
provide for their Subsistence” (1689, First Treatise, §87). Moral elbow room
for this right of individuals to preserve themselves by bringing their self-
owned powers to bear on their extra-personal environment requires that, for
17
Thus, when Locke says that the earth originally “belong[s] to Mankind in common”
he simply means that “no body has originally a private Dominion, exclusive of the rest of
Mankind” (1689, Second Treatise, §26).
214 Eric Mack
each individual, others’ acquisition private property rights and the exercise
of those rights on net leave “enough, and as good” (1689, Second Treatise,
§27) for the use of that individual. It requires that the acquisition and dis-
position of property rights not on net make the extra-personal world less
susceptible to any individual’s efforts to exercise his powers in ways that serve
his ends. So, on this reading of Locke, Locke’s “enough, and as good” proviso
itself arises as a further application of elbow room reasoning. This contrasts
with the Lockean proviso that Nozick endorses which seems to allow acqui-
sitions or deployments of property rights that on net diminish the economic
opportunities of individuals as long as those individuals receive (implicit or
explicit) compensation along some utility or welfare dimension.18
The other example of elbow room reasoning concerns the right of self-
defense. Why is the use of harmful force against another agent morally
permissible when it is employed in order to block that other agent from
violating one’s rights? Why doesn’t the attacker’s right of self- ownership
morally preclude the defender from striking the attacker in order to ward
off the attack? One natural line of argument goes from the prospective vic-
tim’s self-ownership to her possession of a moral liberty to defend herself
against violations of that right and from that moral liberty to harm-inflicting
defensive acts not being boundary-crossings. If individuals are morally at
liberty to defend themselves through the use of harmful force, there must be
moral elbow room for the exercise of that liberty.19 That elbow room can
exist only if agents do not have rights against being subject to such defensive
acts. The elbow room account avoids reliance upon the idea that the attacker
forfeits in part or in whole an original blanket right against being subjected
to any harmful force. This avoidance is a good thing if one holds—as one
should—that self-defensive force may be used against blameless attackers.
For forfeiture can do no real work in explaining the permissibility of those
self-defensive acts. Invocation of forfeiture turns out to be just a back-
handed way of saying that a plausible delineation of rights does not ascribe
to (blameworthy or blameless) attackers a right against being subjected to
defensive force.20
In these cases, as in the elbow room argument for minor intrusions not
counting as boundary- crossings, I have moved from individuals possess-
ing moral liberties to act in certain ways to there being no boundaries
that stand as moral barriers to these actions. However, it might be objected
that such a bold move to the no-boundaries conclusion ignores the possibility
18
On Lockean and/or Nozickian provisos, see Mack (1995).
19
And an individual who is morally at liberty to forcefully defend herself may deputize
another to exercise that liberty on her behalf.
20
Jonathan Quong helpfully pressed me on this issue.
of a more modest conclusion, viz., that there are boundaries that are only
protected by liability rules. I would like there to be a nice, simple, and yet
totally persuasive reason to believe that ceteris paribus an elbow room argu-
ment anchored in A’s moral liberty vis-à-vis B to do x yields the bold con-
clusion that action x crosses no boundary of B rather than the modest
conclusion that B has a right against A doing x albeit a right that is merely
protected by a liability rule. Perhaps that reason is this: to think that A’s
moral liberty is a matter of B having a claim protected by a liability rule
against A’s exercise of that liberty is to make A’s moral liberty too condi-
tional. For, it is to think that A’s moral liberty to x consists in her having that
liberty if and only if she pays a certain price for x-ing.
It is clear in the self-defense case that one gets to the bolder no-boundary
conclusion. The defender does not infringe upon any right of the attacker
that requires compensation on the part of the defender. The defender does
not have to purchase the moral liberty of self-defense from the aggressor.
The same seems true in the case of the moral liberty to use or appropriate
portions of nature. Admittedly, Locke’s argument against the original
joint-ownership of the earth presumes that this ownership would be pro-
tected by property rules—hence, any of the joint-owners may forbid any
use or appropriation of any portion of the earth. Locke does not see the
possibility of and does not refute joint-ownership protected by liability rules.
Nevertheless, as I have suggested, if the moral liberty of preserving oneself
is not itself something one has to pay for, then it seems that elbow room
reasoning that begins with that moral liberty does cut against joint-owner-
ship of nature whether that ownership be construed as protected by prop-
erty or by liability rules.
Still, it is reasonable to ask why moral elbow room reasoning in the minor
intrusion cases bypasses the modest conclusion that those intrusions do cross
boundaries and yet are acceptable if they are accompanied by due compensa-
tion. Why in these cases is it reasonable to go directly to the conclusion that
there is no boundary there to be crossed? I’ve argued that the libertarian the-
orist must reject the general liability rule attenuation of rights and, hence,
must reject the general attenuation route to the permissibility of minor intru-
sions with compensation. However, what is now on the table is the possibility
of basing the conclusion that minor intrusions are permissible on the need for
minor intrusions to be permissible if hog-tying is to be avoided.
From the need for elbow room for minor intrusions, why not draw the
conclusion that, although such intrusions do cross boundaries, these crossings
are permissible with compensation because these boundaries are merely pro-
tected by liability rules?
I think that even in the case of an agent’s moral liberty to do as he sees fit
with his person and property in ways that (non-maliciously and non-wantonly)
216 Eric Mack
involve minor intrusions upon others the consideration that the agent need
not purchase the liberty at issue from those others has force. Consider again the
first case involving A and the over-conscientious B. A’s doing as he sees fit with
his person and property sometimes mildly intrudes upon B. Some smoke from
A’s barbeque lands on B’s acreage and some noise from his motoring up his
driveway vibrates B’s eardrums. However, in contrast to almost anyone else we
might envision, by her choice B never disposes of her person or property in
ways that are even mildly intrusive upon A. Since B does not even mildly
intrude upon A, B does not receive any implicit in-kind compensation in
the form of gains from such intrusions. So, if A’s minor intrusions require
compensation to B, it must be that B can morally require that A desist from use
of his smoker and his driveway unless A makes sufficient explicit payments to
B. That is, it must be that A has to purchase from B his liberty to use his smoker
and driveway.
However, A seems to have a strong elbow room response to B’s demand
that A purchase those liberties from B. The response is that, absent special
circumstances, to have a right over one’s person and property includes hav-
ing the moral liberty to dispose of one’s person and property as one chooses
and that appreciation for the hog-tying problem calls for a recognition that
this moral liberty includes one’s discretionary disposition of one’s person
and property even if that disposition involves minor intrusions on others. A
points out that, in light of what rights are supposed to do—he might say, in
light of the telos of rights—a sensible delineation of rights will not count
minor intrusions as boundary-crossings.
A can further point out that this delineation of rights does not unilater-
ally favor him; B also is credited with the moral liberty to dispose of her
person and property as she sees fit even if her chosen dispositions involve
minor intrusions upon A. Both parties’ rights are construed in a way that
best articulates persons’ underlying moral claim against being precluded
from living their own lives in their own chosen ways. There is, therefore, a
sense—nicely expressed in Bramwell’s live-and-let-live rule—in which B is
implicitly in-kind compensated for A being credited with the moral liberty
to engage in minor intrusions. For B too is credited with that liberty. Note,
however, that under elbow room “compensation,” each individual’s com-
pensation takes the form of the moral liberty to do as he or she sees fit with
his or her person and property without paying compensation even if the exer-
cise of that liberty involves minor intrusions upon others.21
21
Similarly, one could say that each agent is “compensated” for being subject to self-
defensive force by being herself at liberty to engage in such force. However, this is not to
be confused with self-defensive force being permissible only if the attacker is provided
with compensating utility or welfare.
Still, I have left open the possibility that, under special circumstances,
libertarian rights may take the less stringent form of claims that are protected
merely by liability rules. Two quite different sorts of case readily come to
mind. First, there are cases of “soft paternalism.” Suppose B is subjected
without her consent to an interference that would normally count as a rights
violation. It is commonly argued that such interference may be permissible
if B would agree to it in light of the disvalue for B of the outcome that the
interference would prevent. For example, B is unknowingly about to step in
front of a speeding bus. Because of the disvalue for B of being squashed like
a bug, B would waive her right not to be yanked out of the bus’s path.
However, time constraints make it impossible for A to elicit that waiver
from B. Due to those constraints, B cannot exercise choice over whether she
will be yanked back or not. Under such circumstances, it seems pointless to
say that B’s right against such intervention includes a right to determine by
her choice whether A’s interference will be permissible or not. So it seems
that, under such circumstances, B’s right against such interference amounts
to a right not to be subject to it unless she would agree to it were such agree-
ment possible. If this is correct, then under these special circumstances, B’s
right amounts to a right to her being by her own lights duly compensated
for being subject to that interference—and being saved from being squashed
like a bug duly compensates B.22
In a second sort of case, what A has a right to and what B has a right to
are so closely connected that neither party can engage in a core exercise of
his or her right without precluding the other from engaging in a core exercise
of his or her right. For example, A cannot extract his natural gas (which lies
below his plot of land) without precluding B from retaining her natural gas
(which lies below her neighboring land) and vice versa because the gas is
part of a single pool. In situations of this sort the closest we can get to preserv-
ing the right of each to extract his or her own gas and not be subject to the
unchosen extraction of his or her gas is to ascribe to each a moral liberty to
extract gas from the pool subject to a requirement that the extractor compen-
sate the other party for the unavoidable extraction of the other party’s nat-
ural gas; A may extract B’s gas but must pay B the value of B’s extracted gas
minus the costs to A of the extraction (and sale) of B’s gas.23
22
See Steven Wall’s “Self-Ownership and Paternalism,” The Journal of Political
Philosophy, vol. 17, no. 4 (2009), 399–417. Wall argues that advocates of libertarian self-
ownership have no satisfactory explanation for the permissibility of yanking B out of the
path of the bus.
23
The detailed law governing these matters differs from state to state. In Colorado, B
has the option of sharing from the start in the financial risks of the extraction project or
not sharing in that risk and later paying a higher percentage of the costs if the project is
successful.
218 Eric Mack
7. ABSTRACT RIGHTS VS. CONCRETE RULES
There is no philosophical deduction of a bright line between minor and

non-minor intrusions. Even if philosophy could identify something like a
physical scale of intrusiveness; e.g., number of cigarette smoke particles
per cubic foot of air conveyed toward one’s nose or lungs, no point on that
scale would mark the natural boundary between minor and non-minor
smoke intrusion. One reason for this is that whether an act is a minor
intrusion or a non-minor intrusion also depends upon expectations about
what one will and what one will not be subject to. Introducing some
cigarette smoke into a non-consenting person’s lungs went from being a
minor intrusion to an impermissible intrusion partially because more was
learned about the dangerousness of cigarette smoke but also partially
because smokers and non-smokers came more to expect its prohibition.
Also, even if there were a philosophically detectible bright natural line
between minor and non-minor cigarette smoke intrusions, detection of
that line would not provide a great deal of guidance about where the
bright line might be between minor and non-minor noise or odor
intrusions. It is enough for armchair philosophy to say that any concrete
realization of basic rights that is supposed to define for each individual a
domain within which she may do as she sees fit must deem an array of
intrusions to be minor and, hence, permissible.
Still, it is necessary to have some picture of how processes that are not a
matter of armchair philosophizing can give rise to a body of concrete rights
(with respect to intrusions on persons and their just holdings) that would be
acceptable from the perspective of abstract Lockean rights theory. Imagine
an array of judges who each at least implicitly take individuals to have claims
over their persons and their holdings and accept the Bamford v. Turnley dam-
age vs. annoyance distinction. And imagine that, over decades or even gen-
erations, such judges make conscientious decisions in many different sorts of
nuisance and trespass cases.24 They consider cases involving noises, odors,
fumes, vibrations, light transmissions and blockages, and so on. An impor-
tant element in many of these decisions is whether or not the actual conduct
of the defendant should be construed as a boundary-crossing damage or a
24
Judicial decisions are more apt to track Lockean thinking than legislative decisions
because judges are supposed to focus on the rights of the parties who come before them
while legislators are apt to advance policies the rationale for which has little or nothing to
do with rights.
non-crossing annoyance.25 In forming many of their decisions these judges

ask question like these: Do we have here a degree of smoke (or noise or vibra-
tion) intrusiveness, which, if forbidden, would deprive the defendant of his
right reasonably to use his property? Or do we have here a degree of smoke
(or noise or vibration) intrusiveness which is reasonably construed as damag-
ing to the complainant and, hence, as not within the moral liberty of the
defendant? And, as a result of their deliberations, these judges issue not
unreasonable judgments like these: A’s barbecuing in his backyard and A’s
nailing down a new roof for his house are no more than minor (and not
boundary-crossing) intrusions upon neighbor B; but A’s conducting a (hide)
tanning operation or A’s using his backyard as a pile-driving demonstration
site are damaging boundary-crossings for B.
Individuals adjust their expectations and conduct to these not unreason-
able judgments about which actions are damaging and which are (at most)
annoyances because they anticipate other judges reaching similar decisions
in similar subsequent cases; and that anticipation is to some degree self-
fulfilling since judges are likely to reach similar decisions in similar subse-
quent cases because people have adjusted their expectations and conduct on
the basis of that anticipation. If a judge now is confronted with a case not
about the transmission of odor or noise but, rather, vibration, his delibera-
tions will be more complex. For he will not deliberate simply about the
first-order issue of whether the degree of vibration at hand is damaging or
only annoying but also about what judgment in this vibration case is rea-
sonable in light of the lines that have been drawn in prior decisions in odor
and noise cases and in light of the reasonable expectations that individuals
have formed with respect to vibration on the basis of those prior decisions.
And new cases may force judges to rethink and reformulate the principles
that judges previously thought to be embodied in proper decisions; and that
rethinking and reformulation will affect the precedential significance of past
decisions and people’s future expectations.
The result of such processes will be a much more determinate delineation
of certain of persons’ rights than rights theorizing can deliver. Of course,
there is nothing pre-ordained about the resulting (but not permanently
fixed) structure of concrete rights. Many other different structures could
have emerged that would be equally acceptable from the point of view of
25
Other issues would include: (i) whether an apparent trespasser does not intrude at
all because he possesses an easement; (ii) whether the complainant cannot enjoin an
already established activity that damages her when she comes to the neighborhood; and
(iii) whether the dire straits of the defendant provides him with a dispensation to engage
in conduct that would otherwise be boundary-crossing.
220 Eric Mack
abstract rights theory. However, I believe that, if an acceptable structure

actually emerges among us by way of, e.g., the judicial decision-making
processes I have invoked, it (rather than any of the other abstractly equally
acceptable structures) properly governs our interactions. I do not have an
account of why precisely the actually emergent and abstractly acceptable
structure of rights has moral traction for us while the other unrealized struc-
tures do not. However, I think that these sketchy remarks about a body
of articulated rights emerging from the successive not unreasonable deci-
sions of (at least implicitly) Lockean-minded judges is enough to shift the
responsibility for more fine-tuned articulation of persons’ Lockean rights
away from the armchair Lockean.
I conclude, therefore, that Lockean rights theory can do as much as it has
to do to provide an explanation for why minor incursions are permissible—
an explanation that, because it turns on moral elbow room reasoning, does
not open the door to consequentialist reasoning that would be unwelcome
to the advocate of Lockean rights.
Bibliography
Bamford v. Turnley (1862). 122 Eng. Rep. pp. 27–34.
Calabresi, G. and A. Douglas Melamed, A. D. (1972). “Property Rules, Liability,
and Inalienability,” Harvard Law Review 85(6), pp.1089–128.
Cohen, G. A. (1995). Self-Ownership, Freedom, and Equality. Cambridge: Cambridge
University Press.
Epstein, R. (1979). “Nuisance Law: Corrective Justice and its Utilitarian Constraints,”
Journal of Legal Studies 8(1), pp. 49–102.
Epstein, R. (2009). “Property Rights, State of Nature Theory, and Environmental
Protection,” New York University Journal of Law and Liberty 4(1), pp. 1–35.
Filmer, R. (1652). Observations Concerning the Original of Government. In: Patriarcha and
Other Writings, J. P. Sommerville (ed.). Cambridge: Cambridge University Press, 1991.
Gaus, G. (2012). The Order of Public Reason. Cambridge: Cambridge University
Press.
Locke, J. (1689). Two Treatises of Government, Peter Laslett (ed.). Cambridge:
Cambridge University Press, 1960.
Lomasky, L. (1987). Persons, Rights, and the Moral Community. Oxford: Oxford
University Press.
Mack, E. (1995). “The Self-Ownership Proviso: A New and Improved Lockean
Proviso,” Social Philosophy and Policy 12(1), pp. 186–218.
Mack, E. (2010). “The Natural Right of Property,” Social Philosophy and Policy 27(1),
pp. 53–78.
Mack, E. (2012). “Nozickian Arguments for the More-Than-Minimal State.” In:
Bader, R. and Meadowcroft, J. (eds), The Cambridge Companion to Nozick’s
Anarchy, State and Utopia. Cambridge: Cambridge University Press, pp. 89–115.
Mill, J. S. (1859). On Liberty. Indianapolis, IN: Hackett Publishing, 1978.

Nozick, R. (1974). Anarchy, State and Utopia. New York, NY: Basic Books.
Railton, P. (1985). “Locke, Stock, and Peril: Natural Property Rights, Pollution, and
Risk.” In: Gibson, M. (ed.), To Breathe Freely: Risk, Consent, and Air. Totowa, NJ:
Rowman and Littlefield, pp. 89–123.
Schmidtz, D. (2010). “Property and Justice,” Social Philosophy and Policy 27(1),
pp. 79–100.
Sobel, D. (2012). “Backing Away from Libertarian Self-Ownership,” Ethics 123(1),
pp. 32–60.
Wall, S. (2009). “Self-Ownership and Paternalism,” Journal of Political Philosophy
17(4), pp. 399–417.
9
Rules and Rights
Jonathan Quong & Rebecca Stone
This chapter addresses the following two questions that arise when some
agents are imperfectly motivated to comply with the demands of justice or
hold false beliefs that lead them to act unjustly:
• What rules should we devise in response to this non-compliance?
• May such rules license departures from the rights individuals would
possess under conditions of full compliance, and if so, do the resulting
rules actually alter agents’ rights?*
Problems of non-compliance don’t seem particularly troubling for theo-
ries, like utilitarianism, that focus on maximizing the aggregate or average
amount of some agent-neutral conception of goodness. This is because the
problem doesn’t fundamentally change for such theories when we move from
ideal to non-ideal conditions. Our duty is always to maximize the good as the
theory defines it. The particular prescriptions change and will be second best
under non-ideal conditions, but the underlying objective is the same. When
the theory prescribes governance by rules, moreover, agents have only
instrumental reasons to obey those rules; they have a duty to obey them
only insofar as doing so will promote the good.
For deontological theories, by contrast, the nature of the problem seems
to change dramatically in the face of non-compliance. Such theories assume
that we have moral claims against one another that are not reducible to
judgments about how much agent-neutral goodness actions or rules will
produce. Our duties to obey rules that define the scope of our moral claims
against one another are therefore not instrumentally grounded in some
* For many helpful comments and suggestions, thanks to Richard Child, Andrew
Lister, Jeff McMahan, Tom Sinclair, Hillel Steiner, Zofia Stemplowska, Chad Van
Schoelandt, and audiences at the University of Leeds, the University of Manchester, the
University of Wales, Newport, and the Oxford Studies in Political Philosophy Workshop
in Tucson, Arizona.
Rules and Rights 223
independent conception of goodness. But this means that it is not obvi-

ous what we should do when some or all agents fail to comply, beyond
attempting to bring non-compliers into compliance and remedy the con-
sequences of their non-compliance. Yet, while such measures might suffice
in the face of occasional and relatively trivial instances of non-compliance,
they are unlikely to be satisfactory when non-compliance becomes signifi-
cant and widespread.
Political theories that are broadly deontological have thus recently come
under sustained criticism as having little or no practical relevance (e.g.
Farrelly 2007; Sen 2009). Such theories tend to be formulated at the level
of ideal theory, where full, or near full, compliance is assumed.1 But, the
critics point out, we live in a world that is very far from this ideal—a world
where non-compliance is pervasive and its costs are severe. And, therefore,
theories such as Rawls’s have nothing very helpful to tell us about what
rights people possess or how we ought to behave once the assumption of full
compliance is dropped.2 As Colin Farrelly (2007, 853) puts it: “By ignoring
the realities of non-compliance and scarcity of resources, Rawls’ theory of
‘justice as fairness’ insulates itself from the most pressing concerns that face
every real society.”
We argue that this view is mistaken. Deontological theories can inform
the selection of the rules society ought to enforce under conditions of
non-compliance. Although these rules may not perfectly reflect the rights
people have under conditions of full compliance—what we will refer to as
agents’ ideal rights—they may nonetheless be justified from a deontological
standpoint.
And if they are so justified, then agents who comply with them are enti-
tled to demand that others do likewise. And so the rules, though they are
non-ideal rules, alter the content of agents’ rights against one another. This
is not to say that these are the only non-ideal rules that can be justified. Any
deontological theory ought to recognize that there might be weighty
agent-neutral and consequentialist reasons for implementing certain rules
to combat problems of non-compliance that cannot ground moral rights.
We argue that such rules can sometimes be justified all things considered,
but that they do not affect the contours of individual rights. We thus defend
the following two main theses:
(A) Rules designed to address non-compliance can sometimes ground
individual rights. These rights have the same essential features as ideal
rights, and can thus properly be described as demands of justice.
For the classic formulation, see Rawls (1999, 8).

1
Of course, critics of so-called ideal theory have other complaints too, but in this
2
chapter we are focused on the issue of non-compliance.

224 Jonathan Quong & Rebecca Stone
(B) Some rules designed to combat non-compliance may be inconsistent

with deontological principles and as such do not ground individual
rights, even though they are sometimes justified all things considered.
In defending the first thesis, we depart from two other ways of approaching the
problem of non-compliance. According to the remedial approach, the only
rights are ideal rights, and the rules must accordingly respect these rights. Thus,
while the rules may force non-compliers to comply or, when this is not possi-
ble, to make redress, the goal of eliminating non-compliance may not be
accomplished at the expense of the ideal rights of the compliers. In short, as the
name we have given to it suggests, its objectives are purely remedial in nature.
The second approach, rights consequentialism, shares the remedial
approach’s view that the only rights are ideal rights, but departs from it by
denying that the rules must always respect those rights. On this view, the
consequentialist goal of reducing the number of transgressions of agents’
ideal rights justifies rules that permit infringement of them.3
We argue that both these approaches are mistaken. Instead, we suggest
that non-compliance can, at least under some conditions, be viewed as a
burden over which rights can be defined. Deontological conceptions of jus-
tice reflect certain fundamental normative commitments, and these might
be invoked to derive a rights-based solution to the problem of non-compli-
ance, just as they are invoked to determine the content of agents’ moral
entitlements under ideal conditions. The problem of defining rights under
non-ideal conditions is more complex than the problem of defining rights
under ideal conditions. But it is nonetheless amenable to a genuinely deon-
tological solution. This means that the resulting non-ideal rules can consti-
tute a source of genuine moral claims.
Deciding which of these approaches to adopt has important implications
for political and legal philosophy. Many of the important political decisions
and problems we currently face—questions about just conduct in war, the
fairness of affirmative action programs, the design of domestic tax policy, the
design of schemes to address global poverty and inequality, and appropriate
responses to the threat of terrorism, to name only a few—depend in part on
determining how rules ought to respond to the problem of non-compliance
and, insofar as we are deontologists, on deciding whether the resulting rules
generate legitimate individual rights. Yet, explicit discussion of the general
theoretical problem is surprisingly absent from the existing literature.4
3
Following (McMahan 2009, 9–10), we distinguish between rights violations and
rights infringements. The former are unjustifiable transgressions of moral rights, whereas
the latter are justified transgressions of moral rights. We use the term transgression to
cover both types.
4
There is a substantial literature on how consequentialists ought to respond to the
so-called demandingness objection, which the problem of non-compliance allegedly
Before proceeding any further, we should clarify a few key terms and
distinctions. First, we adopt Rawls’s account of the distinction between tel-
eological and deontological theories (Rawls 1999, 21–2, 26). Teleological
theories begin with an independent account of goodness, and then define the
right in terms of maximizing or promoting the good. Deontological theo-
ries reject this picture, either because they do not define rightness in terms
of maximizing or promoting goodness, or else because they do not define
goodness in way that is independent of rightness.5
Second, we use the terms “rights” and “claims of justice” interchangeably to
refer to moral entitlements held by individual persons that are, at least in
principle, enforceable. Our focus is thus on that domain of deontological
morality for which the term justice is usually reserved. We do not consider
whether and what moral duties and claims exist beyond this domain. To act
contrary to an individual’s legitimate rights is thus to act in a way that is,
absent some further justification, unjust by deontological standards. Likewise,
we use the term “rights-based theory” to refer to a theory that aims to provide
a deontological account of individual rights.
Third, for expositional convenience we posit the existence of a rule-
maker who designs rules to respond to the problem of non-compliance.
The rulemaker can be thought of as any decision-making body or process
that makes rules on the community’s behalf; for example, a legislature,
court, or decentralized norm-creating process. We are interested in what
the substantive content of the rules ought to be. We remain agnostic about
questions of institutional provenance and the distribution of political
power in a just society.
The chapter is structured as follows. In section 1 we outline the remedial
and rights consequentialist approaches and argue that they are flawed. In
section 2 we explain the nature of the deontological rulemaker’s decision
problem, and we argue that the costs of non-compliance can often be
treated as a burden over which individual claims can be allocated in a man-
ner consistent with widely accepted deontological principles. Section 3 con-
siders and rejects three objections to our proposal. Sections 2 and 3 thus
constitute our defense of Thesis (A): rules designed to address non-compliance
poses for consequentialist theories, but this is not our focus (e.g., Cullity 2004; Mulgan
2001; Murphy 2000). There also is a growing literature on the distinction between ideal
and non-ideal theory addressing how, if at all, this distinction ought to be drawn and
whether political theorists have been focusing a disproportionate degree of attention on
ideal theory (e.g., Farrelly 2007; Simmons 2010; Stemplowska 2008). But this debate is
also distinct from the question we focus on here.
5
It is important not to conflate the distinction between consequentialism and
non-consequentialism with the distinction between teleological as opposed to deontolog-
ical theories. In this chapter, we remain agnostic as to whether a deontological approach
like ours might be “consequentialized.”
can sometimes ground individual rights. These rights have the same essen-
tial features as ideal rights, and can thus properly be described as demands
of justice. In sections 4 and 5 we move on to explain and defend Thesis (B):
some rules designed to combat non-compliance may be inconsistent with
deontological principles, and as such they do not ground individual rights
but are nonetheless sometimes justified all things considered.
1. THE REMEDIAL AND RIGHTS CONSEQUENTIALIST

APPROACHES TO NON-COMPLIANCE
In this section, we present and reject the remedial and rights consequential-
ist approaches to non-compliance and explain why they are flawed.
1.1 The Remedial Approach

According to the remedial approach, ideal rights are the one and only source
of moral claims, and they set absolute limits on the justifiable content of
institutional rules. Thus, rules that try to solve problems of non-compliance
may not license infringements of these rights. The remedial approach can be
summarized by the following three claims:
(R1) the only moral rights are ideal moral rights;
(R2) justifiable rules do not direct or permit agents to infringe the rights of
those who comply with the demands of justice;
(R3) so long as they satisfy (R2), rules may be designed to direct agents to
respect the ideal rights of other agents, force would-be non-compliers
to comply, and remedy the problems arising from past acts of non-
compliance.6
The problem with the remedial approach is that it doesn’t provide a satisfactory
solution to the problem of non-compliance, at least when non-compliance
is serious and widespread. This is because (R2) forces us to make an unpal-
atable choice: either admit that (R2) prevents the rulemaker from compel-
ling agents to come to the assistance of innocent victims of non-compliance,
thus severely limiting the extent to which problems of non-compliance may
be dealt with; or else construe agents’ ideal rights in such a way that they can
be forced to assist innocent victims of non-compliance without falling afoul
6
Notice that the possibilities set out in (R3) don’t necessarily constitute an exhaustive
list, and so the remedial approach may allow for rules designed with other ends in mind,
so long as they don’t fall afoul of (R2). For example, the remedial approach might permit
rules designed to promote certain forms of human excellence.
of (R2), but at the cost of eviscerating much that is normatively compelling

from the idea of a right.
Consider the latter option first. The extent to which (R2) constrains the
rulemaker’s enforcement and remedial powers depends on the nature of
the ideal rights. More specifically, it depends on whether agents’ ideal
rights are such that others may force them to enforce other agents’ claims.
For if this kind of compulsion doesn’t infringe agents’ rights, then (R2) won’t
inhibit the establishment of effective enforcement schemes.
But how can this kind of compulsion be permitted without eviscerating
what is appealing about the idea of a right? One possibility is that agents
have enforceable ideal duties to aid victims of non-compliance, so that their
ideal rights are not infringed when they are compelled to do so. This looks
like a plausible way out. Indeed, we will ultimately argue that such duties
do arise under non-ideal conditions.
This is not, however, an option for the remedial theorist for three reasons.
First, the remedial approach affirms (R1): there are no moral rights except
ideal rights. Under ideal conditions, there is no basis to infer the existence
of affirmative duties to help victims of non-compliance because, by defini-
tion, there is no problem of non-compliance in an ideal world. It seems
misleading to present a right as “ideal” when the content of the right makes
explicit reference to conditions of non-compliance. Second, the remedial
approach is supposed to capture the intuitive thought that one’s ideal moral
rights cannot be limited or diminished by the unjust behavior of others so
long as one complies with the requirements of justice oneself. To allow for
ideal rights that contain conditional clauses requiring that compliers bear
substantial burdens in the event that others commit injustices thus seems
inconsistent with the underlying spirit of the remedial approach. And third,
even if the remedial approach could coherently incorporate such condi-
tional clauses into agents’ ideal rights, doing so would misrepresent the
nature of the problem of non-compliance. If compliers are forced to assist
the victims of non-compliance, this cannot be what justice ideally requires.
What justice ideally requires is either that no rights violations occur, or else
any victims of rights-violations are fully compensated by those guilty of the
violations. To incorporate within the content of ideal rights conditional
duties on the part of compliers to bear costs created by the injustice of oth-
ers would thus misrepresent as ideal a situation that is clearly non-ideal. If
compliers are (as we will argue) sometimes duty-bound to share in the costs
of non-compliance, this requirement cannot be presented as part of ideal
theory, but must rather be explained in some other way.7
7
A defender of the remedial approach might argue that the victims of non-compli-
ance have an (ideal) right to be given the nearest attainable approximation of resources to
Of course, claim rights are standardly understood to be enforceable:

claim-holders and, possibly, others are at liberty to use force to prevent and
remedy rights violations. That is partly why such rights form the basis of
claims of justice. But while the liberty to enforce one’s rights, for example,
by exercising one’s right of self-defense, affects the claim rights of those
who seek to violate rights, it has no impact on complying agents’ claims
so long as there is no enforceable duty for compliers to aid victims of
non-compliance.8
Thus, the remedial approach, as we understand it, cannot allow for the
possibility that compliers may be under an enforceable duty to aid victims
of non-compliance (though they are, of course, at liberty to do so). This
may sound stipulative, and it is to some extent. We present the remedial
approach simply as one possible way of conceptualizing the relationship
between ideal moral rights and the fact of non-compliance—one that has
been influential within political philosophy. The important point for our
purposes, however, is that, at best, the remedial approach does not point to
an obvious way to solve the problem of non-compliance, since it doesn’t
explain how duties designed to address non-compliance are to be derived.
One of our aims is to think more systematically about this problem.
But what if instead of assuming that agents have enforceable duties to
enforce other agents’ claims, the proponent of the remedial approach
instead construes agents’ claim rights in such a way that they aren’t violated
when they are conscripted into the service of enforcing the rights of others?
If an agent’s right to φ means only that others have a duty not to interfere
with his φ-ing unless they fully compensate him for the interference, then
an agent might be conscripted into action on behalf of others who have had
their rights violated. This is the view of rights adopted by Hillel Steiner
(2006), and accordingly he argues that no injustice occurs when agents are
compelled to enforce others’ claims, so long as they are fully compensated
for their efforts.
There is, however, a problem with such a model of rights. Rights to [φ
or receive adequate compensation] don’t provide agents with the freedom
to choose whether or not to φ. Thus, such rights don’t generate protected
spheres in which agents are free to choose how to live their lives. An
which they are ideally entitled, and that this right explains why there are duties to assist
them. But the relevant question would then be this: who bears the duties that correlate
with this right? As we indicate above, it seems clear the remedial approach must say it is
the non-compliers who bear any such duties. There is no basis, within the terms of the
remedial approach, for assuming that complying agents can be the subject of such duties.
8
Note that this is compatible with the existence of other kinds of enforceable ideal
positive duties—for example, duties to aid victims of misfortune that can’t be attributed
to anyone’s non-compliance.
agent’s freedom can be taken away at any moment by anyone willing to pay
him adequate compensation. Someone might object that this is not so
because an agent has not been adequately compensated unless the amount
he receives in compensation leaves him rationally indifferent between the
option to freely choose whether or not to φ and being forced to φ and
receiving that amount in compensation. But the mere fact the agent is ren-
dered rationally indifferent between these two scenarios by adequately com-
pensating him doesn’t alter the fact that he no longer has a right to φ, but
only a disjunctive right [to φ or be forced not to φ and receive adequate
compensation]. The latter kind of right guarantees him a certain level of
well-being, but it doesn’t confer upon him a genuine freedom of choice
(Coleman and Kraus 1986, 1339–40). He is not free, for example, to insist
on selling his right to φ only for more than the amount of compensation
that would leave him rationally indifferent. Such a refusal might be irra-
tional. But we ordinarily think that rights confer upon agents the freedom
to take actions that are irrational or even wrong (Waldron 1981).9
If, therefore, even adequately compensated conscription is inconsistent
with respecting a plausible conception of agents’ ideal rights, it will be dif-
ficult to devise rules that can effectively combat non-compliance. But with-
out rules that require agents to assist victims of non-compliance, enforcement
will depend on the willingness of agents to volunteer their services to the
project of enforcement, and to the extent that they aren’t sufficiently forth-
coming, the consequences of non-compliance will fall haphazardly on
whomever the non-compliers decide to target—a result that stands in ten-
sion with the deontological commitment to the equality of persons.
1.2 Rights Consequentialism

Rights consequentialism shares with the remedial approach the view that
ideal moral rights are the one and only source of moral claims. However,
unlike the remedial approach, it denies that ideal rights set absolute limits
on the justifiable content of the rules. Rights consequentialism can be sum-
marized by the following claims:
(C1) the only moral rights are ideal rights;
(C2) justifiable rules may direct or permit agents to infringe agents’ ideal
rights;
(C3) such justifiable rules minimize the weighted sum—where weights reflect
each right’s relative importance—of transgressions of agents’ ideal rights.
9
For a detailed critique of the attempt to combine the libertarian commitment to
self-ownership with such a “cross and compensate” model of moral rights, see Sobel (2012).
(C1) is identical to (R1). (C2) is the negation of (R2), and thus captures
the central difference between rights consequentialism and the remedial
approach. (C3) lists the goal of rules that justifiably infringe agents’ ideal
rights. In short, although rights consequentialism views agents’ ideal
rights as the ultimate sources of moral value, it allows, indeed requires,
that the rules be designed to allow the rights of some to be traded off
against the rights of others in order to minimize the weighted sum of
rights transgressions.10
Such an approach has a lot of prima facie plausibility when compared
with the remedial approach. Unlike his remedial counterpart, the rights-
consequentialist rulemaker can forcefully intervene on behalf of the
innocent victims of non-compliance by forcing others to assist in his
enforcement efforts so long as the result is a lower number of rights
transgressions overall.
Notice also that rights consequentialism can be easily modified to accom-
modate a concern with equality. (C3) states that the goal is to minimize
the weighted sum of rights transgressions, but plausible versions of rights
consequentialism can be constructed in which this purely aggregative cri-
terion is replaced or qualified by distributional criteria that require that
rules be designed to ensure that rights transgressions are not too unevenly
distributed.
The problem with rights consequentialism is that it isn’t truly compatible
with the ideals that underpin most deontological theories. Since (C2) per-
mits rights transgressions for the sake of the greater good, rights consequen-
tialist rules don’t generate claim rights of any kind.
Thus, even though (C1) implies that the greater good is measured in
terms of agents’ ideal rights, the resulting rules don’t really respect the rights
on which they are grounded. Agents’ duties to obey the rules are impersonal
or agent-neutral duties that aren’t grounded in particular agents’ claims.
Given widespread non-compliance, the rules will regularly require rights
infringements, with the result that agents will typically have many imper-
sonal duties to infringe the claim rights of others. Thus, it is not clear what
is really left of agents’ claim rights. They continue to make genuine claims
on agents in the interstices of the rules. But when the problems of non-com-
pliance are severe, the freedom that is implied by the interstitial respect of
agents’ claims becomes less and less meaningful. In Nozick’s terms, rights
consequentialism leads to a “utilitarianism of rights” that requires us “to
violate someone’s rights when doing so minimizes the total (weighted)
amount of the violation of rights in society” (Nozick 1974, 28–9). Thus, it
10
Amartya Sen (1982) famously proposed that minimizing rights transgressions could
be one of the goals of a consequentialist theory.
doesn’t give full expression to the deontological insight that individual

rights are not interchangeable units of agent-neutral value.
As we will explain, any deontological theory ought to recognize that
agents’ claims may sometimes be subordinated to the greater good. But
unlike rights-consequentialism, which allows agents’ rights to be sacri-
ficed whenever this facilitates pursuit of the good, we argue that individ-
ual rights must be accorded a degree of priority such that sacrifices of
individual rights for the sake of the good are permitted only in extreme
cases.
2. THE DEONTOLOGICAL SOLUTION
An adequate theory of deontological justice should give proper expression

to principles that reflect a view of persons as free and equal. First, it should
endow every agent with a set of claims and privileges that are compatible
with similar claims and liberties being possessed by all and direct individu-
als or institutions to take appropriate measures to ensure such rights are
respected. That is, agents should be given equal rights, or else any inequali-
ties in agents’ rights must be justified in a manner that coheres with the
moral equality of persons. We will remain neutral among the various prin-
ciples of distributive justice—prioritarian, maximin, sufficientarian, etc.—
that give content to this egalitarian ideal: any plausible distributive principle
grounded in the idea that persons are moral equals qualifies as egalitarian in
the loose sense we intend here.
Second, the theory should accord each agent a set of basic rights and
liberties that have strongly weighted priority relative to other moral claims,
thus creating a protected sphere of individual liberty. In other words, the
theory must require that some individual rights cannot be justly infringed
even when doing so would yield great benefits. This precludes certain forms
of aggregative reasoning that treat individuals as substitutable units of
agent-neutral value, where costs imposed on a person can be outweighed by
commensurate benefits delivered to others.11
We will refer to these principles of equality and liberty as the core deon-
tological principles. But we do not say that these are the only principles of
deontological justice. Nor do we deny that there will be reasonable disagree-
ment about the correct interpretations of these principles.
11
And, as we explain in sections 4 and 5, even if there is sometimes a weighty conse-
quentialist justification for infringing basic individual rights, our view differs from rights
consequentialism in identifying this as a lesser evil justification, and not one that is
grounded in an appropriate conception of justice.
We have just seen that the remedial and rights consequentialist

approaches to the problem of non-compliance fail to properly instantiate
these two principles. The remedial approach either fails to adequately
instantiate the liberty principle by adopting an unduly restrictive concep-
tion of freedom, or else it leaves innocent victims of non-compliance
without adequate mechanisms of redress, thus failing adequately to
instantiate the equality principle. Rights consequentialism plainly disre-
gards the liberty principle, and may not adequately instantiate the equal-
ity principle depending on the precise formulation of the goal that the
rules are designed to serve.
Can we design rules to deal with the problem of non-compliance in a
way that adequately reflects both principles? The central contention of this
chapter is that we can, and that when we succeed in doing so the resulting
rules may endow people with moral rights, even though they may license
departures from compliance with the ideal.
In an ideal world, there is no need for rules to deal with the problem of
non-compliance: justice endows each agent with a set of rights (his ideal
rights); all agents know the contours of other agents’ rights; and all agents
willingly respect those claims. Exactly how justice determines the content of
agents’ ideal rights is obviously a very complex problem, and we don’t tackle
it here. We focus only on the question how they may change when we relax
the assumption of perfect compliance, while remaining agnostic about the
exact content of those rights. We hope, however, that our solution to the
rulemaker’s problem can be rendered consistent with any plausible deonto-
logical conception of the content of agents’ ideal rights.
What happens, then, when we relax the assumption of perfect compli-
ance? Non-compliance can have motivational or epistemic sources: agents
may fail to comply with justice’s demands because they lack the motivation
to comply, or because they hold false moral beliefs or false empirical beliefs
that lead them to act wrongly. Although we have said that, in general, the
remedial approach is inadequate, it may be satisfactory when the number of
instances of non-compliance of any kind is very small. Occasional victims
of non-compliance may be able to adequately protect themselves by relying
on their own defensive resources and the good will of others. But the reme-
dial approach, as we’ve defined it, can’t adequately deal with the problems
posed by widespread non-compliance. Under such conditions, decentralized
and voluntary enforcement efforts will leave many agents unprotected, and
so the rulemaker ought to establish rules to centralize agents’ enforcement
efforts—that is, rules requiring agents to relinquish some of their ideal
rights to the state. Under these conditions, the rulemaker may design rules
that reorganize relationships among agents in the community. That is, the
rules may license actions that infringe agents’ ideal rights.
The distinguishing features of our approach arise from its commitment

to the following two theses:
(A) Rules designed to address non-compliance can sometimes ground
individual rights. These rights have the same essential features as ideal
rights, and can thus properly be described as demands of justice.
(B) Some rules designed to combat non-compliance may be inconsistent
with deontological principles and as such do not ground individual
rights, even though they are sometimes justified all things considered.
(A) is the negation of (R1) and (C1) and represents the central difference
between our approach and the remedial and rights consequentialist approaches.
Non-ideal deontological rules, we will argue, can sometimes alter agents’ moral
rights: it is not true that the only moral rights are ideal rights. However, as
(B) makes clear, we do not deny that there may also be justifiable rules that
don’t alter agents’ moral rights but instead allow agents to infringe the claims
of others. (B) is consistent with rights consequentialism, which naturally allows
for such rules but inconsistent with the remedial approach as we defined it
(although a modified remedial approach might also permit such rules under
exceptional conditions). Exactly when, why, and how certain rules create moral
rights while others are justifiable on other grounds are important questions
that we explore more fully in the following sections. In this section and the
next, we develop and defend Thesis (A). Sections 4 and 5 defend Thesis (B).
Let us assume, then, that non-compliance is sufficiently serious that the
remedial approach proves inadequate. What does our approach recom-
mend? We suggest the rulemaker should initially seek to design rules that
are consistent with the core deontological principles of liberty and equality.
First, respect for agents’ freedom means that the rules must not deprive any
agent of his most basic rights and liberties, and they must deprive agents of
no more freedom (i.e. no more of their ideal entitlements) than is necessary
to effectively combat the problem of non-compliance (assuming that it is
possible for the community to combat the non-compliance at reasonable
overall cost, a point we return to below). Second, the principle of equality
requires that the rules take equally from each citizen enough to effectively
combat the non-compliance, or else that unequal distributions of burdens
can be justified in a manner consistent with the moral equality of persons
(e.g. because they are justifiable to each person including the least-advantaged).
The costs of non-compliance are thereby spread more equally across the
population instead of falling entirely on those whom the non-compliers
choose to victimize. And the rights the state acquires against its citizens are
the same for every citizen.
We contend that even though the situation is less than ideal, when the
rules satisfy these conditions, they alter agents’ moral rights and so make
claims of justice on the members of the community. That is, they set limits
on what the state may do to its members in order to prevent and remedy
rights violations, while also ensuring that members have a duty to other
complying members of the community to contribute to the state’s enforce-
ment efforts up to those limits. This is because they provide agents with a
way of fulfilling their duties to assist victims of non-compliance in an effi-
cient and fair way while preserving a protected sphere for each agent in
which he is free to do as he pleases.
Importantly, we can think of agents’ resulting loss of freedom as some-
thing that has been stolen from them by non-compliers. The non-compliers
ought to return citizens’ stolen rights to them by ceasing their non-compliance
and compensating them for their past acts of non-compliance. But when
the non-compliance is recalcitrant, the community has to prevent the costs
of non-compliance being born arbitrarily by the victims of the non-compliance,
and when it does so in a just way—that is, in a way that respects the free-
dom and equality of persons—new moral rights and duties between the
complying members of the community and between the complying mem-
bers and the state arise.
Is there any room for additional rights-generating rules that might more
effectively tackle the problem of non-compliance? Unlike rights consequen-
tialism, our approach won’t permit the state to take more from certain citizens
simply in order to more efficiently deal with the problem of non-compliance.
Suppose, for example, that non-compliance arises predominantly from a
particular ethnic group in the community, thus making racial profiling an
efficient state response from a rights-consequentialist standpoint. Rights
consequentialism would naturally permit such a policy since it allows non-
compliance to be tackled more effectively at lower cost; that is, members of
the community who were not part of the targeted group could cede fewer
of their ideal rights to the community, and on average the rules would tend to
better preserve agents’ ideal claims.12 Our approach, by contrast, won’t permit
such rules to alter individual moral rights if the only justification for the rule
is its efficiency. Such rules essentially end up using one set of citizens for the
benefit of others, and so fail to respect freedom and equality.
12
More complex versions of rights consequentialism can, of course, take distribu-
tional aims directly into account, and thus might not be as vulnerable to the objection
raised in the main text. But these versions of rights consequentialism still differ from our
own approach in at least two important respects. First, for rights consequentialists, the
resulting rules do not ground individual moral rights—they are simply instruments to
achieve the best possible distribution of ideal rights transgressions. Second, the rights
consequentialist is committed to the proposition that all rights are fungible, whereas our
account is not committed to this idea. We remain open to the possibility (indeed it seems
very likely) that once the principles of liberty and equality are properly specified, it will
turn out that there is no general fungibility of rights.
3. CLARIFICATIONS AND OBJECTIONS
In this section we clarify our approach by considering and responding to

three objections.
3.1 Departures from Ideal Theory Sacrifice Justice

The first objection runs as follows. Since agents’ ideal moral rights perfectly
reflect what justice requires, departures from this standard must, by defini-
tion, involve sacrificing justice for the sake of some other value. As a result,
the rules designed to combat non-compliance cannot represent genuine
claims of justice.
This objection assumes that sacrificing individual entitlements in order
to promote some other value is the only way of departing from the ideal.
But a departure from the ideal might instead reflect a change in the back-
ground empirical conditions that the rules take for granted.13 If a change in
the rules results from a change in the empirical circumstances on which the
theory is premised, deontological principles haven’t been sacrificed for the
sake of competing moral considerations. Rather, the change results from
the fact that the same deontological principles generate different prescrip-
tions under different empirical conditions. The rules that instantiate deon-
tological principles might, for example, look very different under conditions
of significant scarcity than they do under the conditions of moderate scar-
city that ideal theory often assumes. The same is true, we argue, under con-
ditions of non-compliance, because principles of deontological justice
require that the costs of non-compliance must be distributed in accordance
with principles of equality and freedom.
A proponent of the objection might grant the distinction described
in the previous paragraph, but insist that it does not suffice to rebut it.
It may be that we can sometimes make different empirical assumptions about
the background conditions while remaining wholly committed to a deonto-
logical conception of justice, but we cannot always do so. Radically differ-
ent empirical assumptions can make it impossible to design plausible rules
that remain true to the core principles. For example, if we assume scarcity
so extreme that there are insufficient resources to guarantee everyone’s sur-
vival, the resulting rules cannot adequately reflect the deontological principles
13
G. A. Cohen makes a related distinction between applied rules of regulation that
serve justice in particular and applied rules that are impure in that they serve other prin-
ciples alongside justice (2008, chapter 7).
of freedom and equality, since any plausible rule governing such ‘lifeboat’
scenarios will allow some agents to die so that others may live.
But this response does not vindicate the original objection. We don’t
deny that some changes to empirical assumptions may make it impossible
to design rules that effectively instantiate deontological principles. We make
the weaker claim that the costs of non-compliance do not always have this
effect, for sometimes the costs of non-compliance can be distributed in
accordance with deontological principles. Moreover, it will be very difficult
for the proponent of the objection to resist this conclusion once she grants
that not all changes to the empirical assumptions of ideal theory necessi-
tate departures from the deontological perspective on justice. If our critic
grants, for example, that moderate changes to assumptions about scarcity
or human psychology need not threaten the pedigree of the resulting theory,
why should moderate departures from the assumption of full compliance
be any different?
3.2 The Responsibility Objection

Some may object to our claim that the costs of non-compliance are a bur-
den over which rights can be allocated on the grounds that it misconstrues
the nature of non-compliance. The burdens created by non-compliance are
not like some of the other burdens and benefits to which principles of jus-
tice apply, so the objection runs, since they are not brute facts of nature.
Unlike costs created by conditions of scarcity, costs of non-compliance are
costs that particular persons, namely the non-compliers, are morally respon-
sible for. And so justice naturally demands that those persons be the ones
that bear these costs. As David Miller says: “if Bert steals Anne’s money,
justice does not require Charles to right this wrong. . . . Charles commits no
injustice if he fails to secure the return of Ann’s money, because responsibil-
ity for Anne’s loss rests entirely with Bert” (Miller 2011, 239). And so justice
does not demand that the burdens of non-compliance be distributed in a
way that imposes costs on compliers. We might formulate this objection to
our approach as follows:
Responsibility Objection I (ROI): Those who are morally responsible for creating
costs are the only ones who may be justly required to bear those costs.
ROI, however, has counterintuitive implications. Suppose, for example,
that Albert maliciously throws Betty into the water, and the only person
who can rescue her from drowning is Carl. Most people will agree that Carl
must rescue Betty if he only incurs a small cost by doing so, and that this,
moreover, is a duty of justice. ROI denies this conclusion.
Perhaps the responsibility objection ought to be modified as follows:

Responsibility Objection II (ROII): Those who are morally responsible for creating
costs are the only ones who may be justly required to bear those costs, so long as it
is possible for those responsible for creating them to bear all the costs.
In the example above, it is only Carl who can rescue Betty—it is not possible
for Albert to bear the costs of doing so—and so ROII does not preclude
requiring Carl to rescue Betty as a matter of justice. Thus, ROII accommo-
dates our judgment about the initial example by restricting its scope to sit-
uations in which it is possible to impose the costs on those responsible for
creating them. At first glance, this seems like a plausible response. It will
often be possible for those responsible to bear the relevant costs. Those
responsible for creating the costs can often identify themselves and, for
example, compensate the victims of their non-compliance for the injuries
that they have caused.
But remember that we are not arguing that non-compliance alters what
people must ideally do as a matter of justice. Rather, we are arguing that
when faced with recalcitrant non-compliance, the compliers should treat
the costs of non-compliance as a burden to be distributed among them-
selves. Moreover, we do not deny the basic intuition behind the responsibil-
ity objection. The non-compliers remain indebted to the compliers: their
moral duties are unchanged but for the twist that they are owed to all those
who end up sharing the costs of their non-compliance, not just their imme-
diate victims. It is the moral rights of the compliers with respect to each other
(e.g., Betty’s rights with regard to Carl) that change.
ROII also seems to have counterintuitive implications. In our previous
example, it seems implausible to suppose that Carl cannot be required, at the
bar of justice, to rescue Betty even when Albert could do so, but predictably
won’t, particularly when the costs Carl will incur are low and the benefit to
Betty is so great. Suppose Carl expresses unwillingness to rescue Betty, but a
fourth person, Debbie, who is unable to rescue Betty herself, can make the
following threat that will succeed in getting Carl to perform the rescue: she
will take $20 from Carl’s bank account if he doesn’t rescue Betty. Intuitively,
it seems clear that in making this threat, Debbie does not act unjustly, but
rather enforces a claim of justice that Betty has against Carl.
Consider also this modification of the previous example. As before,
Albert has pushed Betty in the water, and she is now drowning. There is also
a second person, Eric, who is drowning, but no one is morally responsible
for the fact Eric is drowning. Albert is able to rescue either Betty or Eric, but
not both of them. Carl is only able to rescue Betty (at little cost).
Suppose that Albert likes Eric, and so has decided to rescue him, and Carl
knows Albert well enough to realize this. There are now three possibilities:
(1) Albert rescues Betty, resulting in Eric’s death; (2) Albert rescues Eric and
Carl does nothing, resulting in Betty’s death; and (3) Albert rescues Eric and
Carl rescues Betty, resulting in no deaths. ROII tells us that because option
(1) is possible, Carl cannot be required at the bar of justice to rescue Betty.
But this seems absurd. Option (1) is much worse than option (3), and Carl
knows that Albert won’t choose it. It’s hard to understand how the existence
of such an option—one that won’t eventuate and is much worse overall—
can explain why justice permits Carl to allow Betty to drown.
Finally, ROII is likely inconsistent with practices and institutions that are
widely accepted as essential in a liberal democratic society. Any such society
will have civil and criminal enforcement systems. These institutions are
invariably funded out of general taxation: it is not simply non-compliers
who fund the police force and the courts; everyone bears the costs of sus-
taining such institutions. It is implausible to hold, as the proponents of the
responsibility objection presumably must, that compliers cannot be required
to support these institutions as a matter of justice whenever it is theoreti-
cally possible to impose the costs of sustaining those institutions on the
non-compliers.14
The proponent of the responsibility objection might protest that the
criminal justice system is in large part designed to deter non-compliance—
that is, to prevent non-compliance before there exist any non-compliers
who could be required to bear the relevant costs. And once the costs of
supporting the criminal justice system are construed in this way, it is con-
sistent with ROII to require everyone to share in the costs of deterrence,
since these are not costs for which anyone is morally responsible.15
This response, however, is problematic. First, although there is much con-
troversy about the underlying justification of the criminal justice system, not
all of the costs associated with the system can be justified in terms of deter-
rence. For example, the reason indigent criminal defendants ought to be
provided with free legal counsel is not explained by an appeal to deterrence,
yet we take this to be a cost created by the fact of non-compliance to which
all citizens can justly be expected to contribute.16 Second, setting the first
14
It is true that non-compliers could often be required to pay court costs, but the
question here is who may justly be required to bear these costs when forcing the non-com-
pliers to bear them is not feasible.
15
Even in an ideal world, some enforcement mechanisms may be needed to solve
assurance problems, in which case a citizen could ideally be required to bear his share of
the associated costs, but we think it is clear that existing enforcement regimes do more
than merely provide assurance.
16
To the extent that a fairer process is also more accurate and so ensures fewer mis-
takes, deterrence objectives may be served by giving indigent defendants a right to coun-
sel. But presumably considerations of fairness rather than efficient deterrence are the
primary reason why justice requires some form of indigent representation.
point aside, it is implausible to draw a bright line between the ex ante costs
of deterring non-compliance and the ex post costs created by non-compli-
ance. The problem is that many of the policies and actions involved in deter-
ring non-compliance occur only once some people have made it clear that
they intend not to comply. Sending police officers to the home of a woman
who credibly reports that she fears her husband may assault her, for example,
is a cost that arises only because the husband is responsible for engaging in
threatening behavior. More generally, the costs involved in protecting peo-
ple’s rights before a right has been violated but after the potential violator has
given the potential victim grounds to be fearful is a significant feature of
contemporary law enforcement. The distinction between costs for which
no one is responsible and costs that some people are morally responsible for
creating thus does not track the distinction between deterrence costs and
post-non-compliance costs.
3.3 The Fairness Objection

The main problem with the responsibility objection, one might think, is
that it depends on a very specific and controversial view about how costs
generally ought to be distributed among persons. But perhaps there is a way
to formulate the worry that retains the intuitive insight of the responsibility
objection, but is not tied to any particular first-order view about the just
distribution of costs. One solution is to appeal to the more general idea of
fairness in the allocation of duties of justice. Miller (2011) and Murphy
(2000), for example, each argue that it would be unfair to require compliers
to “take up the slack” that arises as a result of the non-compliance by some
members of a group under conditions in which a collective duty falls upon
the entire group. Though they focus on cases where some fail to do their fair
share in the performance of a collective duty, we can broaden the scope of
their argument to include cases where costs arise as a result of non-compliance
with an individual duty:
Fairness Objection: When someone fails to comply with some duty she is under, it is
unfair to require other people (who have done nothing wrong) to bear the costs
arising as a result of this non-compliance. It is unfair because it requires some to do
more than their fair share of what justice ideally requires of individuals.
The fairness objection reformulates what may have been the more funda-
mental insight behind the responsibility objection, but the proponent of
the fairness objection (unlike the proponent of the responsibility objection)
need not endorse any particular first-order view about how costs ought to
be distributed among persons. The proponent of the fairness objection sim-
ply insists that whatever the correct account of justice is, once individual
duties have been correctly assigned, it is unfair to expect compliers to take

up the slack created by the non-compliance of others.
Although the fairness objection has some appeal, we believe it fails.17 The
objection focuses on the distinction between compliers and non-compliers—
that is, it focuses on the unfairness in asking compliers to bear more than their
initial share on account of the unwillingness of non-compliers to do their fair
share. But the objection ignores the equally important distinction between
the victims of non-compliance and those who are not victims.18 It is indeed
unfair that non-compliers should be able to shift the cost of their
non-compliance on to those who act as justice requires, and in a world that
was closer to the ideal this would not occur. As we explained when discussing
the remedial approach, in close to ideal conditions only non-compliers
would be forced to bear the costs of non-compliance. But we are considering
scenarios where the remedial approach fails to redress the problems of
non-compliance. Faced with such scenarios, we must confront the following
question: is it consistent with deontological principles to force the victims of
non-compliance to bear all the costs of non-compliance, or should those
costs be treated as a burden to be shared equally across complying members
of society? We do not believe there is any plausible argument that can be
built from principles of equality and freedom that could justify allowing the
unlucky victims of non-compliance to bear all the costs. To do so would
allow the morally arbitrary fact of who happened to be the victim of non-
compliance to determine differences in persons’ access to resources and other
advantages over which people can claim moral entitlements. This is espe-
cially clear in cases where, in the absence of a response from compliers, non-
compliance would impose severe costs on its victims, but those costs could
be averted by imposing relatively small costs on everyone (e.g. by imposing a
small tax on everyone to fund street lighting to deter rapes that dispropor-
tionately target members of certain demographic groups).19
17
Note that, in addition to the objection presented here, many of the concerns pressed
against the responsibility objection also apply to the fairness objection.
18
This forms part of Zofia Stemplowska’s (2014) critique of Miller. Michael Ridge
(2010) also makes this point about Murphy’s argument.
19
Ridge and Stemplowska reach similar conclusions. Both agree that fairness consid-
erations don’t generate decisive reasons to oppose asking compliers to bear some of the
additional costs of non-compliance, and both suggest that fairness may sometimes
require compliers to bear these costs. However, Ridge’s argument has a narrow scope: his
claim is merely about what constitutes a fair distribution of the costs of non-compliance;
he remains agnostic about what compliers may be required or compelled to do, or what
they ought to do all things considered. By contrast, we propose that, under certain con-
ditions, the costs of non-compliance can be treated as a burden over which claims of
justice can be allocated, and so we insist that there are conditions under which compliers
are under an enforceable duty of justice to bear the costs of others’ non-compliance.
Stemplowska’s argument is narrower than ours in a different way. Like us, she focuses on
In sum, none of the three objections considered in this section represent

a serious challenge to:
Thesis (A): Rules designed to address non-compliance can sometimes ground
individual rights. These rights have the same essential features as ideal rights, and
can thus properly be described as demands of justice.
4. RULES BEYOND RIGHTS
We believe that when institutional rules are correctly designed to ensure

that each complying person shares equally in the burden of combating
non-compliance and that this sharing of the burden is also consistent with
the liberty principle, these rules alter agents’ moral rights. But we do not
claim that all justified rules must meet these conditions.
Suppose that: (a) a rule addressing the problem of non-compliance
appears morally justified for weighty consequentialist reasons; (b) the rule
sanctions transgressions of agents’ ideal rights; and yet (c) any feasible alter-
native would make some people at least as badly off as the worst off under
the proposed rule. Under these conditions, we believe that the rule can be
morally justified all things considered, even if the rule is not consistent with
the deontological principles we set out above.20 This supports our second
main claim, Thesis (B), which we mentioned at the outset.
Rules meeting these conditions that also conflict with the core deonto-
logical principles fall into two main categories. First, there are cases where
the number of people who will not comply with the requirements of ideal
theory is so great that it is impossible for the state to effectively combat
non-compliance via standard methods of enforcement (say because the con-
duct is too difficult to detect without extensive and invasive police proce-
dures). Under these conditions, the best response might be to create an
institutional rule governing individual conduct which: (i) would be volun-
tarily complied with by a large number of current non-compliers or could
enforceable duties, but unlike ours, her argument is limited to cases where some people
are under a duty to rescue others in urgent or dire need. She makes no claims about how
the costs of non-compliance ought to be distributed in other situations.
20
Note that our claim is not that rules that meet the three conditions listed above
necessarily conflict with the deontological principles: many such rules can be consistent
with the principles of equality and freedom. Rather, our point is that meeting the condi-
tions set out above is no guarantee that the rule will be consistent with those principles.
Also note that Thesis (B) does not depend on endorsing this particular account. So long
as one accepts that under some conditions impersonal considerations can outweigh or
override individual claims of justice, then one can endorse Thesis (B) even if one rejects
our particular test (e.g., by rejecting the anonymous Pareto test and replacing it with a
more aggregative principle).
be more easily enforced; and (ii) would result in a state of affairs that would
be much better for all compliers than the status quo or any attempt to force
non-compliers into compliance. Then, even though compliers will be vul-
nerable to having their rights infringed by conduct that is licensed by the
rule, they might not be able to declare the rule to be unjustified all things
considered, since they all do much better under the rule than under any
feasible alternative.
But while they may be all things considered justified, such rules do not,
on our view, generate moral rights because they represent a capitulation to
the de facto power of non-compliers to flout the conduct that is ideally
required. Such rules are not designed to ensure that compliers share fairly in
the costs of combating non-compliance taking the behavior of the non-
compliers as a given. Rather, they are designed to get non-compliers to
behave in a way that is less unjust than their current behavior. This differ-
ence of moral aim explains why these kinds of rules do not generate legiti-
mate moral entitlements.
It may help to illustrate this point with a simple example. Suppose there
is a religious society where a substantial percentage of the current popula-
tion believes that female adultery (but not male adultery) is punishable by
death, and all attempts to persuade these people that female adultery is not
a punishable offense, or to coercively prevent them from acting on this
erroneous belief, fail. Thus, the status quo is that many women are mur-
dered on grounds of adultery. But suppose the rulemaker could get the
population to endorse and conform to the following rule: all female adul-
tery is a serious wrong, and punishable by two years in prison, but no greater
punishment is permissible. Let’s suppose that any attempt to legally
entrench a lesser punishment (or to equalize the treatment of men and
women) would be met with fierce resistance and would lead to a return to
the status quo, but the population’s religious beliefs are just flexible enough
to accommodate the idea that two years’ imprisonment is a sufficient pun-
ishment for female adultery. Under these conditions, the rulemaker might
be justified in proposing and implementing this rule, yet it seems clear that
this rule is not of the sort that actually modifies individual moral rights. The
rule might have a sufficient consequentialist justification, but because that
justification involves capitulating to the beliefs of non-compliers rather
than fairly sharing the costs of combating non-compliance, it cannot alter
individual moral rights.21 It is a lesser evil, but choosing the lesser evil is not
21
The legal rules governing conduct in warfare might have the same structure as the
rule in this example. Jeff McMahan (2004, 730–3; 2008; 2009) has suggested that these
rules might have a valid consequentialist justification but don’t reflect persons’ moral
rights, since justice ultimately distinguishes among combatants according to whether
they have justice on their side, whereas the rules of war must capitulate to the reality that
the same as a just distribution of some burden. To borrow an example from

G. A. Cohen (2008, 83–4), if the parents of a kidnapped child decide that
the thing to do is pay the kidnapper the requested ransom, this is not the
just solution, even if it is the lesser evil.
The second main category of justified rules that will not determine indi-
vidual moral rights are rules that are designed to limit the injustice resulting
from badly designed institutions that reliably result in state officials violat-
ing the moral rights of individual citizens. Suppose, for example, that the
police force has developed a culture of contempt towards criminal suspects,
leading them to put pressure on suspects to confess notwithstanding their
right to remain silent, so that the right to remain silent no longer adequately
protects suspects. (We assume, for the sake of argument, that there is such a
right at the level of ideal theory.) The obvious solution to this problem is to
reform police culture and procedures, but if there is no straightforward way
to do this, at least in the short or medium term, it may be appropriate for
the rulemaker to give citizens additional legal rights against the state. Thus,
for example, the rulemaker might require police officers to clearly inform a
suspect of his right to remain silent on the theory that requiring police
officers to provide suspects with this information (assuming such a require-
ment can be effectively enforced) will improve police behavior and enable
suspects to more freely exercise their rights to remain silent.
What is the moral status of this right to a warning? Let’s suppose that
whereas, in the absence of a police culture of contempt towards criminal
suspects, the basic right to silence effectively protects criminal suspects
from the occasional attempts of rogue officials to extract false confessions
without significantly hindering enforcement efforts, but that the right to a
warning, while it offers innocent suspects more protection, also results in
many guilty suspects avoiding conviction on technicalities, thereby under-
mining the state’s enforcement efforts at the expense of innocent victims.
Then, even if the right to silence is a genuine moral right, the right to a
warning may be no more than a prophylactic measure that is to be kept in
place only as long as the conditions of systemic official non-compliance
persist.22 What this category has in common with the previous category is
that departures from ideal moral rights are not justified as mechanisms for
equitably sharing the costs of non-compliance. Rather, the rules are justified
both sets of combatants will typically believe that they are fighting for a just cause, such
that rules of war treating combatants asymmetrically will be doomed to failure.
22
Cf. Maryland v. Shatzer, 559 U.S. 98, 103 (2010) (describing the measures adopted
by the US Supreme Court in Miranda v. Arizona, 384 U.S. 436 (1966), as “prophylactic
measures to protect a suspect’s Fifth Amendment right from the ‘inherently compelling
pressures’ of custodial interrogation” (emphasis added)).
as a kind of compromise. In this case, criminal suspects gain more rights

against the state, at the expense of victims of crime, because the alternative
of more official non-compliance is, on balance, much worse.23 In these ways,
consequentialist considerations may sometimes ground successful justifica-
tions of rules, but they do not determine that part of morality that allocates
moral rights to individuals in a way that respects the freedom and equality
of persons.
5. FURTHER OBJECTIONS
In this section we consider several objections that might be pressed either

against Thesis (B) or our attempt to combine Theses (A) and (B).
5.1 Moral Rights Cannot Be Permissibly Infringed

How can Thesis (B) be consistent with a deontological theory of justice?
Surely such a view must hold that individual moral rights place strict limits
on the design of institutional rules? As Rawls famously tells us: “each person
possesses an inviolability founded on justice that even the welfare of society
as a whole cannot override. For this reason justice denies that the loss of
freedom for some is made right by a greater good shared by others. It does
not allow that the sacrifices imposed on a few are outweighed by the larger
sum of advantages enjoyed by many” (Rawls 1999, 3). Similarly, Nozick tells
us: “individuals have rights, and there are things no person or group may do
to them (without violating their rights). So strong and far-reaching are these
rights that they raise the question of what, if anything, the state and its offi-
cials may do” (Nozick 1974, ix). Critics might thus protest that deontological
justice cannot countenance the transgression of individual rights.
But deontology, as we understand it, is not committed to the implausible
thesis that consequences are irrelevant to moral permissibility. Deontological
theories, recall, are defined as those that reject teleology either because they
do not define rightness in terms of maximizing or promoting goodness, or
else because they do not define goodness in a way that is independent of
rightness. It is thus clearly possible for such theories to allow consequential-
ist considerations to figure in the justification of a given rule.
We also believe that it is clear that individual persons may sometimes
permissibly infringe the moral rights of others, and that there is nothing
23
In the following section we confront the objection that if an institutional rule is
morally justified then the rule must also determine the content of individual moral
rights.
additionally troubling or problematic in assuming that institutional rules

might countenance such infringements. There are a large number of cases
where virtually everyone agrees that one or more innocent individuals may
permissibly be harmed in order to save a larger number of innocent individ-
uals. Philosophers who are inclined to accept some version of deontology
typically invoke distinctions like doing versus allowing harm or intending
versus foreseeing harm to explain when such actions are permissible, but it
is rarely suggested that what makes the harming of innocent persons per-
missible is that the innocent people lack moral rights in the context in ques-
tion. Furthermore, provided one accepts the view that moral rights can
sometimes conflict, any action-guiding moral theory must countenance the
idea that there can be permissible infringements of moral rights.24
5.2 Justified Rules Must Justify Rights

A different objection challenges our view that morally justified institutional
rules don’t necessarily generate moral rights. If an institutional rule purport-
ing to determine persons’ legal claims and duties in some domain is morally
justified, then surely that rule necessarily generates moral rights? If some law,
for example, grants all persons a legal liberty to φ (no duty not to φ), and
we have conclusively determined that the law is morally justified, then
surely the only conclusion to be drawn is that the legal liberty right is also a
moral liberty right? What else could it be since the law that grants it has a
moral justification?25
This objection, however, is flawed. Consider, for example, a deeply racist
society where a black minority is persecuted by a white majority. The racial
prejudice among the white population is sufficiently deep and virulent that
violence is very likely to result whenever institutions try to integrate the two
populations. Suppose that in light of the widespread violence that would
result from having integrated public parks, the best feasible arrangement is to
have a law which segregates parks by race (until the widespread racial preju-
dice can be effectively defused). The fact that this law has a valid moral justi-
fication (the avoidance of large-scale violence) does not, it seems clear, entail
that individual moral rights are really altered by the law. Does a given black
person now lack the moral liberty to enter “whites only” parks? Surely not.
The rule may be justified by appeal to consequentialist considerations
about predicted violence, and the black person might even have an imper-
sonally based moral obligation not to enter the park when his defiance of the
For a defense of the view that rights can conflict, see Quong (2012).
24
Michael Otsuka (2011) advances an argument along these lines in developing an

25
objection to Jeff McMahan’s distinction between the morality of war and the laws of war.
rule would in fact tend to encourage violence. But these sorts of considera-
tions are not the kinds of considerations that can alter moral rights within
the confines of a deontological theory. The rule in question does not instan-
tiate principles of freedom and equality—the costs of non-compliance
might not be equitably distributed and freedom is not adequately pro-
tected—and so even if we believe that the rule has a sufficient justification,
it does not follow that anyone’s moral rights have been altered. This is thus
a case where persons’ moral and legal liberties come into conflict.
In cases like this it makes more sense to say that there is a sufficient jus-
tification for a rule that will result in a number of rights infringements. As
we said before, this is a lesser-evil justification; choosing the least bad
option is not the same as acting in accordance with the demands of justice.
This particular justification grants state officials the permission to prevent
black persons from entering certain parks, but black persons continue to
have the moral liberty to enter these parks: their moral liberty rights remain
unaffected.
In sum, some valid rules may have consequentialist or lesser-evil justifica-
tions, but unless those rules also instantiate the relevant deontological prin-
ciples, they don’t alter anyone’s moral entitlements. Such rules should be
understood as legally licensing certain infringements of moral rights for the
sake of guaranteeing a greater level of protection overall. These rules allow
the non-compliers to gain at the complying agents’ expense, and giving
such a windfall to the non-compliers is a pragmatic response to the problem
of widespread and recalcitrant non-compliance: it is not a just distribution
of the costs of non-compliance among those who comply.
In claiming that rules can sometimes justifiably override or ignore indi-
vidual moral rights, we don’t purport to be saying something that hasn’t
been said by many others. Our point is rather to emphasize the importance
of avoiding two tempting, but seriously mistaken positions. One is the view
that rules that respond to the problem of non-compliance can never alter
the fundamental moral entitlements granted to persons at the level of ideal
theory. The other is the view that justified rules always affect the content of
individual moral rights. Neither of these views is correct. Whether or not a
rule alters moral rights depends on the type of moral justification for the
rule. When the justification is grounded in deontological principles, the
resulting rules do determine moral rights, but rules can be justified in other
ways, in which case they don’t.
5.3 Practically Irrelevant

Finally, someone might object that it doesn’t matter whether institutional
rules designed to deal with non-compliance change persons’ moral rights.
A rule is morally justified when all of its associated features (including the
enforcement of the rule) are morally justified. So once we know that a rule
is morally justified, so the objection runs, there is no point in asking whether
the rule generates moral rights. The answer is practically inert—devoid of
action-guiding implications.
We can see this objection is mistaken, however, once we reflect on the
question how ought we to respond to those who break the rules. Let’s com-
pare the example of the racially segregated parks with the case (described at
the end of section 3.3) where a small tax is imposed to fund better street
lighting to deter would-be rapists. How should a fellow citizen respond to
rule-breakers in each case? Should we respond in the same way to a black
person that we see entering the “whites only” park as we do to a person who
avoids the taxation requirement in the street lighting case? Clearly we ought
to respond to these two rule-breakers very differently. In the segregation
case, it would not be appropriate for bystanders to condemn or sanction the
black person for entering the park (even if state officials may prevent him
from doing so) at least so long as it is apparent that his doing so will not
trigger any violence. But the same is not true of the person who avoids
paying the tax for the additional street lighting. In the latter case, reactive
attitudes of blame and perhaps even sanctions are appropriate responses to
the rule-breaker on the part of all citizens. Thus, determining whether a rule
generates genuine moral rights is not practically inert. When a moral right
has been violated a different response is warranted.
Moreover, compensation is often appropriate when a genuine moral right
has been violated. This is less likely to be true when a rule has been broken
but no moral rights have been violated.26
Both this objection and the previous one misunderstand the relationship
between the moral justification of rules and moral rights. Whether a given
rule has a sufficient moral justification depends on a wide variety of consid-
erations including, perhaps, impersonal values and also, importantly, what
other rules are currently feasible. The justification of moral rights depends
on a narrower set of considerations. Whether a proposed moral right is in
fact a moral right depends on whether the justification for that right correctly
26
May state officials permissibly punish someone who violates the rule racially segre-
gating the parks? There might be extreme cases where the state would be permitted to
imprison such a person for consequentialist reasons (e.g., because failure to do so would
lead to the sort of violent conflicts the rule was intended to avoid). However, it would be
a mistake to equate such imprisonment with “punishment” in the standard sense of that
word, since the standard sense implies the criminal is liable to such punishment (i.e., has
forfeited his right by virtue of a wrongful act), but that is exactly what we are denying
occurs in such a case. So the state might, in extreme circumstances, be morally permitted
to do something to the rule-breaker that looks like punishment, but it would not be
punishment in the standard sense.
instantiates deontological principles of freedom and equality. But there is

more to morality than principles of freedom and equality, and this is why
we cannot simply derive the contours of individual moral rights by looking
to see what rules are morally justified.
6. CONCLUSION
We hope to have shown that deontological theories of justice can tell us

something about what rights people have under conditions of serious and
widespread non-compliance. The world we live in is deeply unjust and the
costs of non-compliance are overwhelming. But the appropriate response to
this fact is neither to throw up our hands in despair and declare that not
much can be done about it (as the remedial approach recommends), nor to
abandon deontological theories of justice (as rights consequentialism rec-
ommends). Instead, deontologists should remain committed to their core
convictions and seek, when possible, to design rules that tackle the prob-
lems of non-compliance in a manner that respects the freedom and equality
of persons. Such rules may not give everyone what they would have under
conditions of full compliance, but they can determine the content of peo-
ple’s moral rights in an unjust world.
Bibliography
Cohen, G. A. (2008). Rescuing Justice and Equality (Cambridge, MA: Harvard
University Press).
Coleman, J. and J. Kraus (1986). “Rethinking the Theory of Legal Rights,” Yale Law
Journal 95, 1335–71.
Cullity, G. (2004). The Moral Demands of Affluence (Oxford: Oxford University Press).
Farrelly, C. (2007). “Justice in Ideal Theory: A Refutation,” Political Studies 55,
844–64.
McMahan, J. (2004). “The Ethics of Killing in War,” Ethics 114, 693–733.
McMahan, J. (2008). “The Morality of War and the Law of War.” In Just and Unjust
Warriors: The Moral and Legal Status of Soldiers, David Rodin and Henry Shue
(eds) (Oxford: Oxford University Press), 19–43.
McMahan, J. (2009). Killing in War (Oxford: Oxford University Press).
Miller, D. (2011). “Taking up the Slack: Responsibility and Justice in Situations of
Partial Compliance.” In Responsibility and Distributive Justice, Carl Knight and
Zofia Stemplowska (eds) (Oxford: Oxford University Press), 230–45.
Mulgan, T. (2001). The Demands of Consequentialism (Oxford: Oxford University
Press).
Murphy, L. (2000). Moral Demands in Nonideal Theory (Oxford: Oxford University

Press).
Nozick, R. (1974). Anarchy, State, and Utopia (Oxford: Basil Blackwell).
Otsuka, M. (2011). “Licensed to Kill,” Analysis 71, 523–32.
Quong, J. (2012). “Rights.” In The Routledge Companion to Social and Political
Philosophy, Gerald Gaus and Fred D’Agostino (eds) (New York, NY: Routledge),
618–28.
Rawls, J. (1999). A Theory of Justice: Revised Edition (Oxford: Oxford University Press).
Ridge, M. (2010). “Fairness and Non-Compliance.” In Partiality and Impartiality:
Morality, Special Relationships, and the Wider World, Brian Feltham and John
Cottingham (eds) (Oxford: Oxford University Press), 194–222.
Sen, A. (1982). “Rights and Agency,” Philosophy & Public Affairs 11, 3–39.
Sen, A. (2009). The Idea of Justice (Cambridge, MA: Harvard University Press).
Simmons, A. J. (2010). “Ideal and Nonideal Theory,” Philosophy & Public Affairs 38,
5–36.
Sobel, D. (2012). “Backing Away from Libertarian Self-Ownership,” Ethics 123,
32–60.
Steiner, H. (2006). “Self-Ownership and Conscription,” in The Egalitarian
Conscience: Essays in Honour of G.A. Cohen, Christine Sypnowich (ed.) (Oxford:
Oxford University Press), 88–101.
Stemplowska, Z. (2008). “What’s Ideal About Ideal Theory?” Social Theory and
Practice 34, 319–40.
Stemplowska, Z. (2014). “Doing More Than One’s Fair Share” (unpublished).
Waldron, J. (1981). “A Right to Do Wrong,” Ethics 92, 21–39.
10
What Is Wrongful Exploitation?
Thomas Christiano
In this chapter I want to offer an account of unfair advantage taking or

wrongful exploitation. Such an account is of great importance in a world like
ours, which is so much organized around voluntary exchange and voluntary
coordination. Some of these agreements and some of this coordination,
though voluntary, are morally suspect. And it is the job of such an account
to illuminate the morality of the suspect transactions, to articulate the
underlying structure in virtue of which the transaction is morally problem-
atic and to pinpoint the element of these problematic transactions that con-
stitute unfair advantage taking. My main focus will be on unfair advantage
taking in agreement making but the account I offer will, I think, cover cases
of unfair advantage taking in other kinds of cases such as in freeriding on the
cooperative activities of others and abuse of trust. These cases have, I think,
a common structure that can be illuminated by a unified account of the
nature of the wrong.
I will argue that the wrong of wrongful exploitation piggybacks on other
wrongs. Wrongful exploitation is not a primitive kind of wronging. Hence,
disagreement on basic moral principles will affect some of the cases one
takes to be genuine cases of wrongful exploitation and which ones one does
not. Nevertheless, I think that the examination of the nature of wrongful
exploitation does help us discriminate to some extent between different
moral theories since some moral theories may have difficulty accounting for
any of our intuitions about wrongful exploitation. The moralized character
helps us understand, I think, the unity in the very heterogeneous set of cases
that come under the rubric of exploitation. Moreover, I argue that wrongful
exploitation is a distinctive kind of wrong even though it does piggyback
on other wrongs. Finally, unfair advantage taking is a very particular kind
of unfairness. The unfairness attaches to the action of the advantage taker.
In this sense, unfair advantage taking is distinct from unfairness in the back-
ground conditions that agreements take place in. The latter notion is a
structural notion that involves a set of facts that give a person a less than
What Is Wrongful Exploitation? 251
equal say in the process of making agreements. Fairness and unfairness in

this sense are terms that apply to a large-scale institutional structure much
as fairness and unfairness in political decision-making do.1 They need not
refer to the particular actions of particular persons or groups of persons.
I will argue that someone can be absolved of the charge of unfairly taking
advantage of another even when the background conditions of agreement
making are unfair to that other person. I will try to elucidate the relation
between these notions later in the chapter.
My main concern here is to illuminate the nature of the wrong in wrong-
ful exploitation. I will proceed on, and attempt to vindicate, the hypothesis
that there is an underlying unity to the concept of wrongful exploitation,
as distinct from mere exploitation (as in exploitation of resources). So, I will
argue for my favored account by means of counterexamples to problematic
analyses of wrongful exploitation and by showing that my account satisfies
some more general desiderata for conceptions of moral wrongs.
I will proceed here dialectically by discussing some paradigm views of
wrongful exploitation and showing why they fall short when we look at
various examples and then I will introduce my own view and defend it by
examining different examples of exploitation and by showing how it satis-
fies some important desiderata.
1. SUBSTANTIVE AND PROCEDURAL ACCOUNTS

OF EXPLOITATION
One of the main debates in conceptions of wrongful exploitation is that

between what might be called substantive conceptions of exploitation and
procedural conceptions of exploitation. The classical natural law tradition
emphasized the substantive dimension of exploitation. These accounts tend
to focus on some inequality in the things exchanged and urge that there is
wrongful exploitation when there is unequal exchange. In part, Marx’s con-
ception of the exploitation of labor seems to be grounded in this idea that
unequal exchange (in Marx’s theory, when the worker receives less in his wage
than the value of his labor) is exploitative. The common law tradition, in
contrast, attempts to emphasize the procedural dimension. The common law
tradition seems to assert that there is wrongful exploitation when there is
some defect, usually of voluntariness, in the process by which the exchange
has come about. I think that one main trouble with conceptions of exploita-
tion is that they tend to come down too hard on one side or another of this
1
See Christiano (2013).
252 Thomas Christiano
divide while some kinds of exploitation seem to be more procedural while

others are more substantive. Conceptions of exploitation tend to accentuate
the insights in one dimension and then miss the insights in the other dimen-
sion. We need an account that crosses this divide even while showing an under-
lying unity to the idea of exploitation. In the next three sections I will discuss
the procedural accounts, the substantive accounts, and hybrid accounts
of wrongful exploitation, and I will argue that they are all problematic.
Following that I will defend my own conception of wrongful exploitation.
1.1 Procedural Accounts

Let us start with the more procedural accounts of exploitation. One analysis
of wrongful exploitation is that:
(1) A exploits B when A benefits from a transaction with B where
B is not participating voluntarily or there is a serious defect in
the voluntariness.
This seems to be the principal analysis in traditional common law doctrine.
This schema is used to interpret the traditional rescue cases of exploitation.
A comes across a stranded motorist (B) in a snowstorm who is threatened
with potentially severe harm or death if he stays out in the storm. A then
exacts an extraordinarily high fee from B and then saves B. The analysis
above suggests that B is not voluntarily participating in the transaction
because there is no decent alternative to A’s proposal or because in these
kinds of cases there is usually a defect in the will of the person in B’s posi-
tion. Another typical example that this is supposed to capture is the case in
which B is woefully uninformed or B is a minor with little deliberative
capacity to understand the proposed agreement.2
In all these cases, the analysis points to a severe defect in the voluntariness
of the agreement that B is entering into. This analysis is puzzling since it
captures the central case of unfair exchange only by immediately extending
the concept too far. Suppose that A comes across B in the same context but
A offers to help B but asks that B pay for the gas. Or take another case in
which a person has contracted a serious illness that must be treated by sur-
gery lest the person die. The surgeon offers to perform the surgery but only
if some reasonable price is paid. Here, too, the person may have little
acceptable alternative to accepting the proposal. This latter case does not
2
See Bigwood (2004) for an account like this. I think related accounts of exploitation can
be found in Wood (1995) in which taking advantage of someone’s vulnerability is exploita-
tive, and in Satz (2010) in which defects in agency are the basis of exploitation. See also
Vrousalis (2013) for an analysis that relies, in a similarly problematic way, on vulnerability.
seem wrongfully exploitative and yet the characterization of the will of B is

broadly the same. He does not have any acceptable alternative. The account
must class what are intuitively clear cases of wrongful exploitation with
intuitively clear cases in which there is no wrongful exploitation, indeed in
which there is no wrong at all.
Absence of an acceptable alternative for one of the parties to the exchange
is not sufficient for exploitation. It is not clear that it is necessary either.
I think intuitively we can say that sweatshop labor is exploitative in many
circumstances, though this intuition must be vindicated. It is not clear that
those who work in sweatshops have no acceptable alternatives; they simply
have alternatives that are quite bad. It is not that the alternative is severe
harm or death. But I think that there are many circumstances in which
intuitively they are wrongfully exploitative. Price gouging also often has
this character.
Now we may reject the account of voluntariness that requires an accept-
able alternative and look only to the quality of will of the person who is
accepting the proposed offer. In the example of the snowstorm above, the
tow truck operator may encounter a deeply flustered stranded motorist who
is in a state of severe anxiety or fear that undermines fully rational deci-
sion-making. We may say that in this case there is a genuine defect in the
will of the motorist. And we may say that when the tow truck driver benefits
from this exchange, we have wrongful exploitation of the motorist by the
tow truck operator.
But a number of problems beset this analysis. One, it cannot discrimi-
nate between the case in which the tow truck operator offers to help the
motorist for a fair price and the case in which the operator offers to help
only for an extraordinary price. The motorist could be equally flustered in
either case. But only the second case can be a clear case of wrongful
exploitation. Two, the view seems to fail to count as exploitation the con-
text in which the operator helps a perfectly cool and collected motorist for
an extraordinary price. After all, a perfectly cool and collected motorist
would be rational in agreeing to an extraordinary price if the alternative
was likely death or severe injury. But this is a fairly clear case of wrongful
exploitation.
So, we can see a general worry about defective voluntariness theories.
They cannot discriminate between clear rescue cases of wrongful exploita-
tion and rescue cases that involve no wrong. In addition, the defective will
view cannot account for some clear cases of wrongful exploitation.
The rescue cases are at the core of the intuitive cases of wrongful exploita-
tion. Any view that cannot account for all these cases must bear a very strong
burden of explanation, which theorists have not offered. I will suggest that
the best account of wrongful exploitation can accommodate these cases.

And it can show what one would have to argue in order to show that these
are not wrongful cases of exploitation.
1.2 Substantive Accounts

Substantive views have, by contrast, a relatively easy time with the rescue
cases. These are the traditional source of examples for this view. Thomas
Aquinas, for instance, argues that exchange ought to be equal and that in
the putatively exploitative rescue case, it is not equal. Of course, it is rarely
clear what the equality in these theories is equality of or what the basic
measure of the value of the goods exchanged is. Aquinas says that a seller
sells a good at the right price to a person in need if the seller sells it at a price
that reflects the cost to the seller.3 Hence, it is the cost to the seller that
determines the value of the good for the purpose of determining when the
exchange is equal. But it is unclear here how the value of the cost is meas-
ured. One might think that Aquinas, and Aristotle before him, think of the
value of the cost in terms of what is necessary for living a good life. But this
is not articulated and it is hard to see how it can be fully worked out. Marx
argues that the worth of a product of labor consists in the socially necessary
labor embodied in the product and that workers are exploited when they do
not receive that same value. The trouble with this kind of account is that
labor is not a homogeneous value. Alan Wertheimer tentatively suggests
that a competitive market price determines the value of an object and that
an exchange that is not at that price is exploitative.4
For these thinkers it is not the utility of the things exchanged that deter-
mines the relevant measure of the value to be equalized. This would gener-
ate a very counterintuitive account in the rescue cases. The marginal utility
of being rescued to the person rescued is likely to be significantly higher
than the marginal utility of the money demanded even at a highly extor-
tionate price. In the rescue cases, only the rescuer would be the exploited
person. This would be a strange result to say the least.
In addition, Aquinas argues that “ . . . if the one man derive a great advan-
tage by becoming possessed of the other man’s property, . . . the latter ought
not to raise the price, because the advantage accruing to the buyer, is not
due to the seller, but to a circumstance affecting the buyer.”5 Hence, in the
3
See Aquinas (2003), ST II-II Q. 77, Art 1 for the principle of equality in exchange.
4
See Marx (2003), p. 392 for an account of unequal exchange between capitalist and
worker based on the labor theory of value. More recently, see Roemer (1985). And see
Wertheimer (1996).
5
See Aquinas (2003).
rescue case, Aquinas would say that the rescued person’s life is not given to
him by the rescuer but is saved due to the special circumstance the rescued
person is in. The rescuer did undergo some cost in helping the rescued per-
son and that is what the rescuer can be said to give the rescued person but the
life or even the great utility that the rescued person received was not given to
him by the rescuer. A higher price than the market price for the cost of the
rescue is greater than the value that the rescuer contributed.
This is an intriguing argument and there is something right about it but
the basic problem with it seems to be that the value involved in the cost is
itself a function of the “circumstances affecting the buyers.” It is a function
in part of the supply and demand for the object. So, it is hard to see how the
market price is giving us a valuation of the product that is independent of
the circumstances of the buyer.
Once we reject the objective measures of value involved in the Aristotelian
and labor theory approaches, the competitive market price seems like a good
candidate for measuring value. But if we take the competitive market price as
the measure of value, there are a host of difficulties to be worried about. One,
it is often hard to know what market we are talking about. Does it have to be
world market prices or local prices? Is price discrimination generally wrong?
What of the market of like objects? Is that relevant? Two, there often are no
competitive markets for the goods involved. The creation of international
treaties is sometimes said to be exploitative but it is hard to characterize it
in terms of markets. Three, sometimes competitive market prices can be
exploitative as well. Intuitively, at least, Marx is on solid ground when he
asserts that the competitive market price for labor can be exploitative even
when there are a number of employers, as long as there are a lot of unem-
ployed people, the workers are not organized and the alternative to employ-
ment for most workers is very bad, while the alternative to not employing a
worker for an employer need not be bad since there are many others.
The price, supposing the groups of employers and employees are suffi-
ciently well segregated (due to sharp differences in education, wealth, etc.),
will be quite low and intuitively it would seem that it could be quite wrong-
fully exploitative.6 And sometimes the very idea of offering a price for a
certain kind of good suggests exploitation, as in the case of prostitution. We
needn’t agree with every one of these cases to see that the idea of exploita-
tion as unequal exchange measured in terms of market price is not giving us
a necessary condition of exploitation.
There are three other classes of cases that might suggest a more general
criticism of the substantive views. The first is the case of irresponsible
6
See Arneson (1996). See also Roemer (1985), p. 35. Wertheimer acknowledges some
of these difficulties, which is why he is tentative about his proposal.
exchange. Suppose one person wants to buy a car but is not willing to put
any time into determining what a good car is or what an appropriate price
is and so goes into a store willing to get the first car at the first announced
price and does so. Now suppose that the usual job of the car salesman is to
bargain, fully expecting that the buyer will haggle and the price will come
down. Instead, the buyer accepts the first announced price and the deal is
made right away. The salesman even says that they can talk more about the
price but the buyer just waves him off. Here, the buyer has bought a car at
a price that is greater than the market price or the cost of getting rid of the
car or any objective price that might be thought reasonable. But it is not at
all clear that the buyer has been wrongfully exploited even though the deal
may be unequal in some important way. Here, it seems that features of the
process are driving our judgment that the seller is not taking unfair advantage.
The buyer has every opportunity to make a good deal and does not avail
himself of them.
If irresponsible or heedless participation in exchange can legitimize an
exchange even though the exchange is unequal, it appears that substantive
unfairness in the outcome is not a sufficient condition of wrongful exploitation.
This is because a feature of the process, the irresponsible action of the buyer
and the reasonably conscientious actions of the seller, can defeat the idea that
there is unfairness in the agreement even if the price is obviously problematic
by normal market standards. Unequal exchange seems insufficient to justify a
judgment of exploitation.
It seems unnecessary as well in many circumstances. For instance, if
the salesman manipulates the buyer into buying a car at a good price and
the buyer did not want to do this, it seems like there may be exploitation
here as well even though the buyer paid a good price. Here, process seems
to be driving our judgment that there is wrongful exploitation despite
equal exchange.
Two other examples suggest that unfairness in the outcome is not neces-
sary to exploitation either. First, some kind of deception or emotional
manipulation to get someone to engage in an exchange that they do not
want can undermine an exchange even if it is for a fair price. Suppose that
B does not want to sell her car to anyone. But B loves A though A does not
love B. A gets B to sell her car to him, for the market price (or any price that
seems fair), by saying how much he needs the car and intimating that he
might love B if she sells him the car. Or A relies on the existence of some
kind of emotional quirk in B to get B to sell him the car at the fair price.
B becomes flustered under certain circumstances at which point she can be
manipulated into doing something she later regrets. A knows this and
pounces and thereby gets B to sell at the market price. Another case is one
in which A enters into some kind of convoluted exchange with B which has
the effect that B sells her car (without entirely understanding this). She sells
it at the fair market price. But I think we would still want to say that A has
exploited B in these cases.
To be sure, we might be able to apply Aquinas’s idea here that the seller
may sell an item at a higher price if the seller values the item at that higher
price. But even here we may run into trouble. For suppose that B is willing
to sell her car at some price but not to A. A very much wants B’s car. But
A nevertheless manages to get B to sell her car at the price she wants to
sell it through one of the machinations mentioned above. It still seems
that A has exploited B. Now, there may be a further measure of the value
of B’s car that somehow includes the value of selling it to one person rather
than another or on one day rather than another or any set of details that
determine the circumstances under which B is willing to sell her car. But
the standard of value for determining equal exchange seems to become
elusive in these circumstances. For any particular standard it would seem we
can set up an example that suggests that there can be exploitative exchange
even while there is equality. Here, it seems that we have a case in which
we can have an exploitative agreement even though the price paid for the
item is a fair price.
The substantive outcome of the transaction is fair in whatever sense one
wants to think of it but the transaction is exploitative. To be sure, these are
cases of manipulation but they are also cases of exploitation, since one per-
son is benefitting at another’s expense.
Another example of exploitation that Alan Wertheimer discusses seems
to be a counterexample to the thesis that unfairness in the distribution of
goods that arises from an exchange is necessary to an exploitation claim.
The example is one of a psychotherapist engaging in sexual relations with a
client.7 One might think that the psychotherapist is exploiting the client in
this context but it is hard to see that this is the consequence of an improper
distribution of the benefits of the relationship. No doubt there is a rule
with some moral force against sexual relations between therapist and cli-
ent, which is there to protect the integrity of the therapeutic process and
thus in some sense to protect the interests of the client and the therapist.
The therapist violates that rule by proposing sexual relations with the cli-
ent. The therapist is exploiting the client because she is relying on a weak-
ness in the client, which the therapeutic process is designed to help remedy,
to acquire some benefit for herself.
So, against the substantive accounts we can say that they fail to provide
necessary conditions of wrongful exploitation because some cases of
exploitation seem to involve exchange at fair prices, however one wishes
7
See Wertheimer (1996), Ch. 6.
to define them; and they fail to provide sufficient conditions of wrongful

exploitation because some cases of exchange at clearly unfair prices seem
not to be wrongfully exploitative. And against the procedural accounts
offered so far, which rely on some kind of involuntariness, they fail to pro-
vide sufficient conditions of wrongful exploitation since some cases that are
counted as involuntary are not exploitative. And they fail to provide neces-
sary conditions.
1.3 Hybrid Accounts

Some have proposed accounts of exploitation that involve both the proce-
dural and the substantive dimensions of exploitation. For example, Mikhail
Valdman defends the view that A wrongly exploits B when and only when
A extracts excessive benefits from B who cannot, or cannot reasonably,
refuse A’s offer.8 Here, we can see the two elements described above con-
joined. First, the exploitation requires an excessive transfer from B to A.
And second, it must be because B is in a position in which he cannot rea-
sonably refuse the agreement. Valdman does not describe the second condi-
tion in terms of involuntariness; but that is no matter since he is invoking a
condition that many would describe as involuntary, namely the absence of
an acceptable alternative. What is important for us is that he is concerned
with an agreement between A and B in which A benefits excessively and in
which B has no acceptable alternative.9
It seems to me that some of the above counterexamples may not damage
this account. For instance, suppose that A is a surgeon and B is a patient
with a life-threatening cancer that can be removed by surgery. But A only
performs this surgery for a price that is significantly above the market price
for reasons we need not explore. A informs B that there are other doctors
who charge less but B waves him off and says that he wants A to get started.
A performs the surgery for the very high price. Here, it is plausible to think
that the price is excessively high but that B has an acceptable alternative to
which he refuses to have recourse.
But the hybrid view must be subject to counterexamples to it stating
necessary conditions, since there are counterexamples to each of the two
conditions as necessary conditions. For example, A is a surgeon who charges
the right price, whatever that is, but B doesn’t really want A to do the sur-
gery, for some reason. A, knowing this, lies to B in telling B that there are
8
See Valdman (2009), pp. 3, 9. Another account that is close to this is in Arneson
(1982).
9
Valdman characterizes an unacceptable alternative as one in which urgent needs are
not met or in which it is not possible to lead a minimally decent life.
no acceptable alternatives, even though there are. A is simply trying to

increase his client base. B thinks he has no acceptable alternative and so pays
the right price for the surgery to A. It seems to me that A is exploiting B in
this context even though he is not receiving excessive benefits from B and B
has not gotten an offer he can’t refuse.
The problem here is that in the understandable theoretical concern to
understand how exploitation can occur in mutually advantageous and con-
sensual circumstances some theorists forget that exploitation also occurs
under other conditions such as lying, manipulation, and coercion. Once we
broaden the scope of our analysis to include these kinds of cases, we can see
that the above conditions are not necessary.
It also seems to me that the therapist case above is also a counterexample
to the thesis that excessive benefits and lack of an acceptable alternative are
necessary conditions for exploitation. It is hard to describe the therapist as
getting excessive benefits. The problem is that he is getting the wrong kind
of benefit and it is not obvious that the clients have no acceptable alterna-
tive. We have some rather complex moral norms relating to sex. And they
are relevant to the claim of exploitation. For instance, suppose that it costs
a tow truck driver a certain amount to save a person in a storm. We may
think that it is not wrong for the tow truck driver to charge this fee in a
particular instance. But now suppose that the going rate for sexual services
is below this fee and the tow truck driver asks for the relevant sexual service
from the person she is saving along with no further fee. This still seems
wrongfully exploitative even though it is hard to say that the transfer is one
of excessive benefits. I think this judgment holds even if the exploited per-
son is normally a sex worker. The problem is that this is the wrong kind
of benefit.
Another worrisome kind of case is the case of sweatshop labor. It may be
that the employers are receiving too much in this kind of case, but it is not
obvious that the employee’s urgent needs are at risk or even that the
employee will not be able to live a minimally decent human life without the
job or that the job makes the difference between a minimally decent life and
a less than minimally decent life.
What I think is hard to question is that the joint condition above is suf-
ficient for wrongful exploitation. Valdman argues that there is an obligation
not to charge excessive benefits when a person has no acceptable alternative
and this does seem right and may explain the wrongness of exploitation in
the cases that exemplify these features. One possible counterexample to this
may be a case in which the only available surgeon stupidly or unknowingly
or forgetfully charges too high a price for a life-saving surgery. It is not obvi-
ous to me that this is exploitation though something wrong has occurred here.
Another worry is that much will depend on what the notion of excessive
amounts to. We have seen above that this is hard to pin down. My thought
is that we have to import another idea here to explain it, but I will discuss
this in what follows. In any case, since the view does not seem to capture
necessary conditions, it does not give us an account of what exploitation is
or even what makes exploitation wrongful more generally.
2. BEYOND PROCEDURE AND SUBSTANCE
What we have here is a dialectic between procedural and substantive

accounts of exploitation. The procedural accounts seem to fail to give an
adequate characterization of the wrong in the rescue cases because they
falsely imply that nearby permissible cases are also wrongful. They also
falsely assert that a defect in voluntariness must accompany all wrongful
exploitation. The substantive accounts seem to mischaracterize wrongful
exploitation because they imply the wrongfulness of unequal exchange
when that is the result of the irresponsibility of the purported victim.
They also falsely assert that every case of equal exchange is not wrongfully
exploitative. Yet, both of the types of views bring insight to the discussion.
They do seem to capture parts of what we have in mind by exploitation.
But the parts are not individually sufficient and they are not individually
or jointly necessary. And we have also seen that a combination of the ele-
ments brings added insight; it may give us a sufficient condition (though
not always), but it nevertheless fails to offer a set of necessary conditions.
What I want to propose is a conception of exploitation that is neither
procedural nor substantive but that abstracts from these two dimensions
while explaining how the different dimensions have the moral significance
that they do have.
I want to propose a conception of exploitation that is able to account
for at least most of the main cases or at least the principal reasoning that
goes on in the main cases. I am going by the hypothesis that there is a
sufficient amount of unity in the idea of the wrong of exploitation and
unfair advantage taking to make sense of this project despite the fact that
the two rival approaches on offer so far fail to capture the whole of what
is going on in the notion of exploitation. This is a hypothesis that we might
have to reject eventually, but for the moment, I think it can be for the most
part vindicated.
There are some other desiderata that I am looking to satisfy along with
the capacity to account for our intuitive reactions. And these conditions
have to do with the fact that I am pursuing the idea that there is a distinct
notion of wrongful exploitation that has moral content and import. One
condition is that the account of the concept of exploitation that we are
elaborating can explain the idea that the exploiter wrongs the exploitee. It is
not merely the case that something bad or problematic has happened; a
wrong has been done to the exploitee.10 A second related issue is to be able
to explain the specific character of the wrong of exploitation. We will see
that this is an issue for the view I will defend. A third desideratum is the
avoidance of what Pogge calls the moral counterproductivity of such a prin-
ciple.11 In particular, I am concerned with avoiding a principle that would
prohibit opportunities to act that make everyone better off while permitting
people to avoid acting in a way that makes everyone better off. The usual
accounts of exploitation seem to imply this because they suggest that some
mutually advantageous and consensual agreements may be wrongful
exploitation and thereby prohibited. But they also suggest that the exploiter
may withdraw from agreements altogether in preference to a non-exploita-
tive agreement. The consequence of this kind of approach may be that
many persons are made worse off if people act in accordance with their
moral duties but do not do much to act morally beyond what they are
required to do. For instance, it may turn out that many people will be left
unsaved if there is a requirement that the savers do so for free or very little
recompense and there is a permission not to save or put oneself into the
position of being able to save another.
This might seem to protect a person who is being saved from extortionate
demands, but it may also create incentives in people to avoid situations in
which they can help. While if one can demand anything one wants, more
people may put themselves in the position of helping others and thereby
more people may be helped. It is an aim of my theorizing to avoid the two
possibilities of saying that exploitative action is pro tanto permissible on the
one hand or on the other hand saying the exploitative action is impermissi-
ble but one may avoid engaging with the other altogether, thus ensuring the
other is worse off.12
My concern here is grounded in my general sense that moral principles
are generally guided by a concern to advance the interests of human beings.
But I think one may be concerned to avoid the counterproductivity in this
case by a less controversial concern, which is that a principle that is meant
10
I think this is a worry for Steiner (1984).
11
See Pogge (2009), p. 115 and Wertheimer (2010).
12
We will see that the avoidance of counterproductivity in moral principles is quite
different from acceptance of what Wertheimer calls the “non-worseness” principle. This
principle says that if B consents to, and is made better off by, transacting with A then the
transaction cannot be morally worse than A’s not transacting with B. So, if A is permitted
not to transact with B, then any mutually advantageous and consensual transaction is
permissible. See Wertheimer (1996), p. 289. Like Wertheimer, I find the non-worseness
principle quite unintuitive but it is asserted by Zwolinski (2007), p. 708.
to protect the interests of a person not be such that it actually has a ten-
dency to set back the interests of those persons. Avoidance of this feature is
on a par, I think, with avoidance of the implication of leveling down from
a principle of equality.13
I will be content if it can be shown that a principle that prohibits unfair
advantage taking does not, through a combination of requirements not to
exploit and permissions not to help, normally set back the interests of those
it is supposed to be protecting. There may be particular cases in which
everyone’s interests are set back by the application of the principle, but if
these are the exceptions rather than the norm, that may be sufficient to
vindicate the principle under discussion. The moral status of exploitation is
that it is pro tanto wrong, though it may under some circumstances be all
things considered permissible. Indeed, this condition is one that I think
offers support for the view that I will defend in what follows.
3. EXPLOITATION
We can now offer an account of wrongful exploitation or unfair advantage

taking. A takes unfair advantage of B if and only if three conditions hold:
(1) A benefits at B’s expense (2) from B doing various things, and (3) A does
this by violating a duty to B. This is what explains the idea that unfair
advantage taking is a wrong to the person who is taken advantage of.
Let me explain the elements of this account. Condition 1 states that A
benefits at B’s expense. This expression needs to be understood carefully. It
is meant to be compatible with mutually advantageous transactions. In the
case of mutually advantageous transactions, A’s benefit at B’s expense is
meant to refer to the size of the part of the surplus A is receiving from the
transaction. A benefits at B’s expense in the sense that the size of the part
of the surplus B receives could be larger. In the wrongfully exploitative
rescue case, A charges B a very large sum in return for A saving B’s life. B
is still a gainer here, but A benefits at B’s expense. Of course, all exchange
has this feature; it involves transfer of something from B to A. Obviously,
this is not sufficient for wrongful exploitation; the other two conditions are
necessary still.
When we are not dealing with exchange, A’s benefitting at B’s expense
need not involve a transaction of this sort. For instance, someone who films
another person in severe distress and then sells the film, without the subject’s
13
I have tried to show that the best understanding of the principle of equality in dis-
tribution of goods implies that the principle of equality is opposed to leveling down. See
Christiano and Braynen (2008).
consent, is also benefitting at the distressed person’s expense in the sense

that that person’s distress is being broadcast to the embarrassment or humil-
iation of that person. And if A freerides on B’s productive activity, A can be
benefitting at B’s expense. But if A lies to B to get B to take medicine that
benefits B, and this benefits A as well since now B is calmer, there may be a
violation of duty here but it is not wrongful exploitation.14
The second condition is that B does something that makes for this trans-
fer from B to A. B actively participates in the process by which A is benefit-
ted at B’s expense. This condition is meant to rule out the case in which A
kills B in order to collect the inheritance from B or in order to get B out of
the way of some project of A’s. Wrong as this is, it is not intuitively a case of
wrongful exploitation. But there are some difficult limit cases of wrongful
exploitation where B participates only in a very limited sense.
For example, A can wrongfully exploit B by filming B in a state of severe
distress in order to make money. Here, B is participating but only in a rather
limited sense. Even more remote is if A films B after B has been murdered
and A tries to make money off the lurid scene.
The third condition is that A benefits at B’s expense by violating a duty to B.
I will spend more time articulating this condition in the following section, but
the idea can be seen in the rescue case. In the case between two ordinary per-
sons, A has a duty to save B at low cost.15 A violates the duty to B by failing to
save B when it doesn’t cost A much. But A also violates this duty to B by mak-
ing his saving B conditional on B paying much more than the low cost. There
are many other ways in which A can violate his duty to B, but this is sufficient
for the case at hand. If A insists on a high payment from B before saving B,
then he has gotten B to benefit A in a way that violates the duty to B.
It is this condition that enables us to grasp the heterogeneity of the
wrongs of exploitation in both the procedural and the substantive cases.
This condition will generate the different results in the rescue cases since
rescuers do have duties to help and they violate those duties when they ask
for too much in return and they act in accordance with those duties when
they do not. Furthermore, ordinary rescuers will have different duties from
professional rescuers. The latter make their livelihood from the activity and
must be on call, so they may charge higher prices. In the cases of agreements
between persons that are distinguished by great asymmetries of information
and/or some kind of monopoly power, duties limit what people can ask for
their services because the ordinary price mechanisms of markets do not
I thank Cynthia Stark for challenging me with this example.

14
There is no circularity here. The low cost is specified independently of the transac-
15
tion. It is an admittedly vague but very important component of the duty of rescue. See
also Snyder (2013).
work effectively in these contexts and threaten inefficient and inequitable

outcomes.16 I will explore this heterogeneity more in the next section.
A couple of things about this account should be noted here. First, unfair-
ness in the circumstances of agreement making is not sufficient for unfair
advantage taking.
Unfair advantage taking involves something further. It involves how one
responds to unfair circumstances or at least to circumstances in which one has
some kind of duty to remedy the unfair circumstances. Principles of fair
agreements specify an ideal of the circumstances in which agreement making
takes place; principles of unfair advantage taking specify how an individual
must act to deal with non-ideal circumstances in which fairness in the circum-
stances is not fully achieved. They presuppose that one has some kind of duty
of remedy towards the person who is in unfairly unequal circumstances.
And it is usually the case that one does not have a duty to remedy all of
the unfairness in the making of an agreement. So, one may make an agree-
ment under unfair circumstances with another without taking unfair
advantage of that other. Two, one’s conception of unfair advantage taking
will depend on what duties one thinks persons hold towards one another.
For example, I have argued elsewhere that equality of opportunity and
equal resources for understanding agreements are among the basic princi-
ples of fair exchange.17 And I have asserted that each person owes a fair
contribution to the equality of opportunity of all other persons. The duty of
remedy is determined by that basic notion. So, on the view I have charac-
terized, I argue that unfair advantage taking consists in failing to act in
accord with one’s duty of remedy in the context of making an agreement
with a person with lesser opportunities. Alternatively, one might think that
sweatshop labor is exploitative because the employers come from countries
that have wronged or are wronging the workers in the particular part of the
developing world at issue and the employers have some duty to contribute
to rectifying the wrong. Someone who rejects both of these duties may have
a different view of sweatshop labor.
4. SOME EXAMPLES OF UNFAIR ADVANTAGE TAKING

AND HOW TO AVOID IT
Here, I will explain this account by laying out some of the main examples
of exploitation.
16
See, for example, Mill (2008), Book V, Ch. XI for a discussion of the proper extent
and limits of laissez-faire in markets.
17
See Christiano (2013).
4.1 Rescue
In the standard rescue cases, one person (A) is confronted with another
person (B) who needs to be rescued or face a very high risk of severe harm
or death. And by hypothesis, B has no further option because no one aside
from A can help him and because he needs help immediately. What I want
to say is that in the normal case this implies that in the context, A may not
morally withdraw from the interaction unless there is a great cost to him. A
has a duty to rescue B at fairly low cost to himself. This makes it impermis-
sible morally for A to bargain to full advantage with B over the conditions
under which the rescue is to take place. To be sure, A should receive some
fair recompense for helping B, if the costs are substantial. But A must figure
out how to determine fair recompense without hard bargaining with B.
If A does bargain hard with B under these circumstances, A is in effect
saying to B that he will not help B unless B pays a sum that is greater than
necessary to assure A that his cost is low. If A succeeds in this, A is exploiting
B. But he does this by violating the duty to B.
The above case is a case of the ordinary duty to help others at low cost to
oneself. But there are other cases of helping others that do not require such
a low cost. In the case of persons who help others as professionals, such as
doctors or lawyers or tow truck drivers, these may in fact charge a higher
price for their services since their livelihoods and expertise depend on such
payments. There are duties limiting what a professional can ask in these
circumstances but the duty is not limited to making sure that the profes-
sional suffers a low cost. Indeed, it is important that the professional be a
beneficiary in these kinds of cases so that there are adequate incentives for
people to engage in this kind of work. I think these kinds of cases motivate
the fair price norm of exchange and may be the motivations for the substan-
tive views. The reason why is that these cases involve a division of labor with
a great deal of asymmetry of information between client and professional
and they often involve a kind of monopoly power. A society benefits greatly
by the existence of this division of labor and wants to ensure that there are
enough people who will occupy these positions in the division of labor.
Hence, it is important for a society to provide adequate compensation for
these services. Yet, because of the asymmetries of information, it is hard for
the client to determine what the appropriate price is supposed to be. The
client is in a potentially very vulnerable position in need of these services.
Hence, there is a need to have a sense of a fair price to be paid for the ser-
vices. This price may be defined in terms of a kind of market price deter-
mined by an independent appraiser (as in the case of houses) or as
determined by a professional association. But one must note that important
as these cases are, they are not the only kind of case and so the appearance
of a general need for a conception of equal exchange is illusory.
It is worth pausing here to note two things. First, exploitation in the

relation between an ordinary rescuer and a rescued person need not be
exploitation when the rescuer is a professional exercising her professional
skills. And this is because of the difference in the duties we think these dif-
ferent kinds of people have, which in turn may be justified by different
institutional needs. In the case of the professional, it is desirable that she
charge more because we want there to be an adequate amount of such pro-
fessionals on call. This suggests a difficulty in the natural law approach that
insists on evaluating the exchange merely in terms of the values of the things
exchanged. Second, in the professional rescue case, it is perfectly acceptable
that the professional benefit from helping the vulnerable person, and there
is no sense that exploitation is taking place. So, we must reject Nicholas
Vrousalis’ conception of exploitation, which he says occurs when A instru-
mentalizes B’s vulnerability to extract a net benefit from it.18 This is exactly
what the tow truck driver, the doctor, and the salvage ship operator do
when they save people as means to earning their livelihoods.
Valdman considers a view like the one I have sketched here and argues
against it. He considers a case where it would initially cost a particular per-
son A a lot more than he is duty bound to pay in helping someone in need.
He claims that in this case, A is not bound to help B (because the cost to
him is too high). Yet, he argues plausibly, A could help B for an excessive
price. In doing so, he would be wrongfully exploiting B. Valdman takes this
to be an argument against the idea that the duty A has to help persons in
dire need is what explains the presence of wrongful exploitation.19
But it is not obvious that Valdman has understood the moral structure of
the case properly. For if A can be assured of compensation for helping B up
to the point where the aid is a low cost to A and A is the only person who
can help B, then it seems to me that A does have a duty to aid B. And it
seems to me that the price he may legitimately demand is what will ensure
that the cost of his aid is low. Here, it seems to me that the duty to aid
explains the wrongfulness of the exploitation.20
18
See Vrousalis (2013), p. 132.
19
See Valdman (2009), p. 4. Valdman also complains that the duty to aid account
cannot explain why a very excessive price is worse than a somewhat excessive price since
both demands would violate the duty to aid. But I don’t see why we can’t say that the duty
to aid is more seriously violated the more excessive the price.
20
Another objection that I have not had enough time to consider so far is from A. J. Julius:
“A threat to do what’s independently permissible can count as wrongfully coercive. I should
not announce that unless you do my chores I’ll sell a couch of mine that holds fond
memories for you. My broadcast of an intention to report the affair, when it figures in a
plan for manipulating the adulterer’s conduct, seems wrong even on the hypothesis that
I owe such a report to the spouse.” My inclination is to think that the moral environment
is more complicated than Julius is suggesting. We have moral reasons to take the person’s
fond memories into account and not merely to use them. See Julius (2013).
One virtue of the account I am offering is that it seems to me that this

particular account of the wrongfulness of exploitation also implies that A
may not permissibly simply walk away once he learns he may not exploit
B. While the account I am advancing does not permit leveling down, it
also does not suggest the compromise position that sometimes exploitation
is okay.21
The duty to aid creates a kind of baseline for each person to contribute to
helping others in dire straits. One cannot simply avoid the cost by avoiding
people because one has a kind of imperfect duty to help others generally.
This imperfect duty takes on a perfect and directed form when one encounters
a person in dire straits and no one else can help. In essence, I think this is
the appropriate response to Pogge’s worry that morality permits one to
avoid people if one can’t exploit them and thereby ensures that people may
actually be worse off. If there is a general and imperfect duty, one cannot
simply avoid the cost. To be sure, one might still avoid people who are
potentially in need and this may make some particular persons worse off,
but if people act generally in accordance with their general imperfect duties,
these instances will be relatively few.
Further, I want to argue that one cannot simply discharge one’s imperfect
duty to help others and then fail to help a particular person in dire straits
on the grounds that one has already helped. I think intuitively this is the
way we understand these duties. But there is a theoretical explanation as
well. Failing to help a particular person in dire straits when one is the only
one who can help is a public and clear expression of indifference to human-
ity, which is a greater violation than merely failing to contribute to a collec-
tive fund.
But there are puzzles. What if B cannot pay even close to a fair cost either
before or after the rescue? Does this allow A to refuse help? This does suggest
a genuine moral problem but it is not a problem with the analysis of
exploitation.
4.2 Cases of Unequal Opportunity

In the case of unequal opportunity, one party (C) with many valuable
opportunities confronts another (D) who has relatively few opportunities
with not very much value. This implies that D will have less of a say than C
in shaping the agreement between C and D. I think this is the kind of case
we confront when wealthy or relatively wealthy persons engage in economic
cooperation with poor persons with limited means and education. This is
also the kind of situation that people worry about when they are thinking
of sweatshop labor.
21
As Alan Wertheimer suggests in Wertheimer (2010).
Some examples of this are as follows. First, wealthy companies contract-

ing with poor persons in the developing world for labor is a key instance of
this. In these cases, the wealthy companies have a lot of different opportu-
nities and the individuals in the developing world have lesser opportunities
because they are poor and not very well educated and so their skill set is
limited. It is also because the immigration laws of other countries prohibit
them from entering another country for better opportunities. Second,
wealthy persons in the developed world employing migrants from the
developing world are also often instances of very different opportunity sets.
And, of course, affluent persons from affluent backgrounds who employ
persons from poor backgrounds with little opportunity for education are
another main kind of case of this sort.
The moral background behind these cases makes a big difference to our
attitudes towards this kind of case. Here is one kind of principle that would
be relevant to thinking about this kind of case. If we proceed on the basis of
a principle of robust equality of opportunity and argue that all human
beings ought to have such opportunities, then we can argue that all human
beings have duties to contribute some fair share to the establishment of
equal opportunities for others.22 In particular, the better off have duties to
contribute a fair share to the opportunities of the worse off. Hence, the
company that is employing workers in the developing world comes to the
situation with some prior duties towards those individuals. Those duties do
not require the company to rectify all of the inequality of opportunity in the
instance, but only to contribute a fair share to the rectification. Hence, they
owe the developing-world workers something more than the best deal they
can extract from those workers. They owe them something that in effect
contributes a fair share to improving the worker’s opportunities over and
above what they could extract by taking maximum advantage.
So, first, C has a duty not to press for advantage to the maximum. And
C has a duty to contribute to D’s opportunities. To be sure, a company is
not to be compared with individual persons in terms of opportunities. The
relevant comparison is between individual members of the company and
others. But there may be a sense in which the duties of the individual mem-
bers of the company aggregate to the duty of the company. This could
involve any number of things. It might involve C enhancing D’s education
in some way. It might involve C ensuring that D has more of an equal say in
the making of the labor contract. C does not have a duty to rectify the
whole inequality (assuming it hasn’t created it) but just to do its bit in
rectifying it.
22
I defend this principle of equality of opportunity as a principle of fair exchange: see
Christiano (2013).
We shouldn’t think that C is the only participant in this relationship.

After all, there are constraints on C that derive from the fact that C is com-
peting with other firms to sell products to developed-world distributors and
ultimately to developed-world consumers. All of these persons are partici-
pating in the relationship with D and, on the principle I am asserting here,
have duties to contribute to the opportunities of developing-world workers.
Hence, all of these should participate in ensuring better conditions for the
workers who are producing the consumer products. So, not only should C
make some contribution, its consumers should as well, by paying higher
prices. Once we widen the lens to include all the different participants in
the relationship, we generate significant duties to D.
I think this captures much of the discomfort people experience with what
is called sweatshop labor. The trouble is that while companies often treat the
situation as one where the ordinary norms of bargaining and market behav-
ior are legitimate, in fact the ordinary norms of bargaining must be signifi-
cantly curtailed. And this is because of the moral duties of the wealthy
regarding the worst off. The situation is a morally fraught one.
Another potential source of discomfort with the sweatshop labor cases arises
when it is the case that the wealthy company or individual shares in some
responsibility for the unequal opportunities of the developing-world workers.
This may be because of a past history of imperialism or neo-imperialism on
the part of the society of which the company is a part.23 It may also hold
if one thinks that the immigration policies of the society from which the
employers hale are unjust and part of the cause of the low wages. In these
kinds of cases too, the employer and the other participants in the production
chain as well as the consumers have some duty of repair, which they violate
when they treat the situation as one of an ordinary market transaction.
Of course, the responsibilities are more extensive if one holds all these
views.24
The point here is that C and all of the beneficiaries in the chain of produc-
tion and consumption exploit D by engaging in ordinary market behavior
and hard bargaining (which might be quite legitimate among persons who
have equal opportunities) and thereby benefitting more from the transaction
than is permissible.
Notice that even if C transacts with D in a non-exploitative way, this will
usually not eliminate all the unfairness in the background situation between
For a powerful example of this kind of view, see Pogge (2002).

23
I have not yet arrived at an account of exactly how to determine how much the
24
employer might owe in any of these circumstances. My guess is that while the employers
usually owe significantly more than the market wage is in these cases, there is a significant
amount of uncertainty here and that there is no exact wage or set of working conditions
owed. This would seem to conform to the view that I am defending.
C and D. Much background unfairness may still remain. But in doing its
fair share in alleviating this unfairness C is engaging in non-exploitative
behavior even if unfairness in the circumstances remains.25 Here, we see
how the key distinction between the idea of background or structural
unfairness and unfair advantage taking makes a difference to our discussion.
But there may still be some discomfort here. For what is required to defeat
a charge of wrongful exploitation on my analysis here is not that one pro-
vide the worker with some particular amount of good, say a living wage, but
that one do one’s part in helping the worker. What this will amount to
cannot be determined in advance and may sometimes fall short of some-
thing like a living wage.
Here too one might worry that this account will discourage people or
groups like C from employing people like D because we have argued that C
must carry a heavier cost than a simple market interaction would require.
This would be a worrisome implication.
D would plead with C to employ him even at the exploitative wage. D
would be worse off if C were to attempt to avoid the transaction (as would
C). Again, we would have something like a leveling down worry here. The
very moral principles that are designed to help the vulnerable worker would
actually damage the worker. And I think that many accounts of exploitation
either imply this leveling down or end up accepting in some grudging way
the permissibility of exploitation.
But here too the account I am offering gives us a more satisfying approach
to these problems. I want to argue that the duty in virtue of which C’s mar-
ket behavior towards D is exploitative is one that C has independent of the
market transaction. C cannot escape the cost by simply avoiding the devel-
oping world. In effect, there is a cost that C must carry no matter what C
does and so this cost should be understood to be the moral baseline in terms
of which C ought to be reasoning about the case. C cannot say, “In order to
avoid the cost of non exploitative transactions I will simply avoid transac-
tion with D.” This is because C must suffer this cost anyway as a matter of
moral requirement.
To be sure, it is permissible for C to avoid this particular transaction, but
the view I am trying to defend here suggests that C will not benefit from
25
Hence, I think Alan Wertheimer’s worry that it would be “unreasonable to expect
the better-off party to repair those background conditions [of social injustice] by adjust-
ing the terms of a particular transaction” can be assuaged once we see that the duty of
repair is not to solve the whole problem but to contribute one’s fair share to the solution.
This can amount to quite a bit when we consider all the participants in the chain of
production but it is not going to be unreasonably demanding on any particular partici-
pant. See Wertheimer (1996), p. 234. See also Young (2004) for a nuanced discussion of
the responsibilities developed-country persons have to developing-world persons.
avoiding the transaction because C (and all the other wealthy participants
in the supply chain including the consumers) have imperfect duties to help
so that they do not avoid the costs by avoiding the transactions. This is a
way in which the view I am defending here avoids counterproductivity
since avoidance of transactions do not benefit persons. At the same time,
the view I am defending rejects the “non-worseness” principle, which asserts
that a mutually beneficial and consensual transaction cannot be morally
worse than the avoidance of transaction. In any case, C may permissibly
avoid the transaction with D, but C cannot avoid the cost.
One last note to this kind of case: C may not discharge its duty of doing
its fair share of enhancing the opportunities of human beings generally in
the world (by say contributing to a fund that does this) and then exploit the
hell out of D in particular. This is for the same reason as we saw in the duty-
of-aid case. There is something special about the public “in your face” char-
acter of exploitation and the failure to do one’s duty to this particular person
that makes this action more problematic than mere failure to contribute to
the opportunities of people generally.
4.3 Cases of Diminished Capacity

Another potential site of unfair advantage taking is one in which E con-
fronts F who has diminished capacity due to mental disability or to a poor
education or poor information.
For the sake of ease of exposition let us assume first that these flaws are
faultlessly possessed by the vulnerable party. In one kind of case, F simply
has a diminished capacity or little information or understanding of what is
involved in the making of the agreement, while E has a great deal of under-
standing, capacity, and information. F cannot figure out what the agreement
actually consists in, while E can. In another kind of case, the vulnerable party
F has a lesser capacity to determine the terms of the agreement in a way that
advances the concerns F has or in a way that advances F’s interests. At the
same time E, by hypothesis, does have this ability.
There is evident unfairness in contexts like these. And I think the unfair-
ness, once again, can be understood in terms of a kind of inequality of say
in the process of constructing the agreement. In the first case, F may be
benighted or confused as to what the content of the agreement is and how
to shape the contract in one way or another. The most obvious and extreme
kind of case like this is one in which F does not know the language in which
the agreement is expressed. Less extreme cases involve technical terms F
does not understand or the agreement may be so complex that F does not
have the time to figure it out. In these kinds of cases, F simply has a dimin-
ished capacity to shape the contract. The second kind of case involves F not
having the knowledge to understand what the implications of the agree-

ment are. As a consequence, E is likely to have a much more efficacious say
over the content of the agreement than F.
Assuming F’s ignorance or incapacity is faultless, E has a duty also not to
press for full advantage in this context but to give assistance to F in terms of
information and cognitive resources. Here, too, the duties people have are
complex and vary from context to context. In the case of the division of
labor, professionals have duties to explain the complex transactions to ordi-
nary persons who are engaged in them. And we structure by law many trans-
actions where a great deal of information not normally possessed by ordinary
persons is grasped by someone who has expertise in this activity. Persons
have rights to lawyers when they are accused of crimes. People usually employ
agents when they buy houses and houses must be legally appraised for their
value. Doctors must abide by the requirements of informed consent when
they recommend treatment to patients.
Violation of these norms, when it redounds to the advantage of the pro-
fessional, the seller is often exploitative.
5. FAIRNESS AND THE LEVELING DOWN PUZZLE
There is a worry with any such set of requirements that they may have the
perverse effect of dissuading the wealthy from employing the worse off.
This could have the perverse effect of making the worse off even worse off
than they would be were they to be taken advantage of. This does seem
like a legitimate worry. It is somewhat analogous to a leveling down
worry.
My response to this invokes the definition of unfair advantage taking,
which includes a violation of a prior duty. The idea is that the wealthy can-
not avoid the duties of contributing their fair shares to the worse off by
avoiding employment relations. They still owe the duties. And they owe
them to everyone. So the fair share they would be required to pay would be
required anyway. In effect, it would determine a baseline. To say that one
does not want to employ very poor persons on the ground that one does not
want to discharge the duty of repair and thereby avoid the charge of unfair
advantage taking is confused. One owes the duty anyway. The question is
how one discharges it.
But this does suggest another puzzle. Could the wealthy pay a sum
towards some development agency and then take full advantage of the
worse-off persons in labor contracts? It isn’t clear this would make any sense
economically but in any case it seems problematic. The reason why this is
problematic is because taking advantage of the unequal opportunities of
(say) developing-world workers is in some way worse than merely failing
to discharge the duty of opportunity enhancement. Taking advantage of
unfortunate workers is a kind of public and “in your face” way of treating
persons as less than equal. It is highly expressive of a failure to recognize the
worker as an equal.
Failing to discharge one’s duty more generally, since it is not directed at
any particular person, does not have the same public meaning.
6. THE DISTINCTIVE CHARACTER OF THE

WRONGFULNESS OF EXPLOITATION
I have argued that wrongful exploitation can only be explained with the help
of a prior duty to a person that is violated, which duty is independent of and
prior to the duty not to exploit. But I think there is, nevertheless, a distinc-
tive duty not to exploit. This is because exploitation introduces a distinctive
dimension to wronging a person. It in some crucial way seems perversely to
make the person a part of the wronging. It makes the wronged person a
participant at least in the sense that the person is acting in a way that makes
exploitation possible. In contrast, think of wrongfully benefitting from your
wrong to another person that is not exploitative. If A kills his parent so as to
receive the inheritance from the parent, A is benefitting from wronging the
parent. But A is not exploiting the parent because the parent’s actions do not
currently contribute to the illicit advantage A gains. In contrast, A’s exploita-
tion of B involves A’s deriving a benefit from B’s activities. B is involved. This
seems to make it a more intimate kind of wrong to B. It seems an even more
public and clear way in which B’s interests are subordinated to A.26
7. CONCLUSION
I have defended a conception of wrongful exploitation according to which

A exploits B when and only when A derives an advantage from the actions
of B and A derives this advantage because A has violated a duty towards B.
I have tried to illustrate how this account works and how it can explain the
plausible parts of the procedural and substantive conceptions of exploita-
26
See Snyder (2013).
tion. I have also tried to show how the account can avoid the problem of
counterproductivity or leveling down that seems to attend many views of
wrongful exploitation. I take this to be a major point in favor of the view.27
Bibliography
Aquinas, Thomas (2003). On Law, Morality and Politics. Ed. Richard J. Regan and
William P. Baumgarth. Indianapolis: Hackett.
Arneson, Richard (1982). “What’s Wrong With Exploitation?” Ethics (91) 2,
pp. 202–27.
Arneson, Richard (1996). Book Review of Exploitation, by Alan Wertheimer, Mind
(110) 439: pp. 888–91.
Bigwood, Rick (2004). Exploitative Contracts. Oxford: Oxford University Press.
Christiano, Thomas (2013). “Equality, Fairness and Agreements,” Journal of Social
Philosophy: Special Issue on New Directions in Egalitarianism, Vol. 44, n. 4 (Winter
2013): pp. 1–22.
Christiano, Thomas and Braynen, Will (2008). “Inequality, Injustice and Leveling
Down,” Ratio, Vol. XXI, n. 4 (December).
Julius, A. J. (2013). “The Possibility of Exchange,” Politics, Philosophy and Economics,
Vol. 12, n. 4 (November).
Marx, Karl (2003). Capital: A Critique of Political Economy in The History of Economic
Thought: A Reader. Ed. Steven Medema and Warren Samuels. London: Routledge.
Mill, John Stuart (2008). Principles of Political Economy. Ed. Jonathan Riley. Oxford:
Oxford University Press.
Pogge, Thomas (2002). World Poverty and Human Rights. London: Polity Press.
Pogge, Thomas (2009). “Testing Our Drugs on the Poor Abroad,” in Exploitation
and Developing Countries. Ed. Jennifer Hawkins and Ezekiel Emanuel. Princeton,
NJ: Princeton University Press.
Roemer, John (1985). “Should Marxists Be Interested in Exploitation?” Philosophy
and Public Affairs, Vol. 14, n. 1 (Winter), pp. 30–65.
Satz, Debra (2010). Why Some Things Should Not Be for Sale: The Moral Limits of
Markets. Oxford: Oxford University Press.
Snyder, J. (2013). “Demeaning Choices,” Politics, Philosophy and Economics, Vol. 12,
n. 4: 313–58.
Steiner, Hillel (1984). “A Liberal Theory of Exploitation,” Ethics Vol. 94, pp. 225–41
Valdman, Mikhail (2009). “A Theory of Wrongful Exploitation,” Philosopher’s
Imprint, Vol. 9, n. 6 (July).
27
I would like to thank the participants in the Oxford Studies in Political Philosophy
Conference in Tucson, October 2013, the participants in the Rocky Mountain Ethics
Conference in Boulder, Colorado, August 2012, and the participants in my seminar on
Exploitation in the Spring of 2013 for helpful comments. In particular, I thank Violeta
Ignieska, Stefan Sciaraffa, David Estlund, Cynthia Stark, Alistair Norcross, Michael
Tooley, David Schmidtz, Houston Smit, Meena Krishnamirthy, Andrew Williams, and
an anonymous referee for OSPP for valuable discussion on the chapter.
Vrousalis, Nicholas (2013). “Exploitation, Vulnerability and Social Domination,”

Philosophy and Public Affairs, Vol. 41, n. 2 (Summer), pp. 131–57.
Wertheimer, Alan (1996). Exploitation. Princeton: Princeton University Press.
Wertheimer, Alan (2010). “Exploitation and Clinical Research,” in Rethinking the
Ethics of Clinical Research: Widening the Lens. Oxford: Oxford University Press.
Wood, Allen (1995). “Exploitation,” Social Philosophy and Policy 12.
Young, Iris Marion (2004). “Responsibility and Global Labor Justice,” Journal of
Political Philosophy, Vol. 12 (4), pp. 365–88.
Zwolinski, Matt (2007). “Sweatshops, Choice and Exploitation,” Business Ethics
Quarterly 17 (4), pp. 689–727.
PA RT I V
METHOD
11
Value-freeness and Value-neutrality in
the Analysis of Political Concepts
Ian Carter
Political philosophers attempt to give coherent accounts of power, freedom,

justice, rights, democracy, and other concepts that feature in our discussions
about the nature, explanation, and desirability of certain political phenomena.*
In doing so, they engage in a kind of conceptual analysis. What makes such
an analysis a good one? What are the criteria, for example, for a good analysis
of power? Some such criteria are fairly uncontroversial. We should seek an
account that will fit, at least to some degree, with common-sense claims
about power (for example, that Barack Obama is a very powerful man). We
should try, moreover, to unpack the notion of power in a way that is logically
consistent in terms of the relations between different kinds of power (such
as coercion and manipulation) and the relations between power and other
phenomena (such as the authority of certain powerful people or the
unfreedom of those subject to their power).
A more controversial methodological question concerns the role of ethical
evaluations in the analysis of political concepts. When we describe certain
people as under the power of others, we seem to be condemning their situ-
ation in ethical terms. Similarly, when we say that one society is freer than
another, we seem to be making an ethical comparison in favor of the former.
Should we say, then, that an adequate analysis or definition of power or
freedom must presuppose certain ethical evaluations? In other words,
should we think of power and freedom as ethically loaded concepts? Or
should we say, on the contrary, that the best accounts of power or freedom
* I am grateful to Boudewijn de Bruin, Keith Dowding, Matthew Kramer, Elijah

Millgram, Peter Morriss, Pamela Pansardi, Mario Ricciardi, Hillel Steiner, Peter
Vallentyne, Albert Weale, and two referees (one of whom was David Schmidtz), for some
very helpful written comments on earlier drafts. I am also grateful for the extensive feed-
back I received both at the Oxford Studies workshop in Tucson and at previous seminars
at the universities of Bayreuth, Braga, Budapest, Manchester, Oxford, Pavia, and Virginia.
280 Ian Carter
allow us simply to describe how things are in the world, without presuppos-
ing or implying any ethical judgments, so that power and freedom qualify
as “value-free” or “value-neutral” concepts? Do power and freedom have
negative and positive value in the same ways as those in which, say, injustice
and justice do? If so, it seems that to describe an interpersonal relation as
one of power is necessarily to disapprove of it (at least pro tanto), and to
describe a situation as one of freedom is necessarily to approve of it (at least
pro tanto). Or is it the case that we first describe a particular situation as one
of power or freedom, and then evaluate it ethically? The latter view seems
plausible in light of the fact that a supporter of the American Constitution
might, after all, judge the power exercised by Barack Obama to be ethically
permissible or required, and an Islamic fundamentalist might, after all, judge
the less free society (on a liberal definition of freedom) to be ethically supe-
rior. It matters which of these views is correct, for the answer will affect how
we understand the very nature of phenomena like power and freedom and,
in turn, the ways in which we justify certain evaluations or prescriptions.
In this chapter I shall defend the use of value-free and value-neutral con-
cepts in political philosophy. As I see it, some reasons for working with such
concepts have been overlooked, owing to a failure to make certain relevant
distinctions. This failure has brought about a misleading polarization of the
debate: on the one side, there are those who affirm the possibility and use-
fulness of defining certain political concepts in a way that is wholly detached
from any ethical concerns; on the other side, there are those who deny the
usefulness or even the possibility of a value-free or value-neutral analysis of
any of the concepts that figure in our normative political discourse.
The position I shall defend is more complex, in two senses. First, I shall
propose a distinction between value-freeness and value-neutrality (I use the
slightly strange-sounding noun “value-freeness,” in preference to “value-
freedom,” to avoid confusion with freedom understood as a political value).
Value-freeness concerns the place (or rather, the absence) of evaluative terms
in the definition of a concept, whereas value-neutrality involves a suspen-
sion of judgment about the comparative merits of divergent ethical posi-
tions. Each of these two features is a methodological desideratum in the case
of certain concepts in certain theoretical contexts. Sometimes, it is appro-
priate to adopt an empirical approach to the analysis and definition of a
concept, and this requires a value-free definition. Sometimes, it is appropri-
ate to adopt an ethically non-committal approach, and this requires a degree
of value-neutrality. The two approaches can be combined, but, as we shall
see, they can also come apart.
Second, I shall propose a distinction between value-freeness and value-
neutrality, on the one hand, and the complete detachment of our analysis
from all ethical concerns, on the other. I shall call the latter idea “value-
Value-freeness & Value-neutrality 281
independence.” One can have ethical reasons for adopting an empirical

approach to the analysis and definition of a given concept. And one can
have ethical reasons for adopting an approach that is non-committal with
respect to a given set of ethical positions. Thus, neither value-freeness nor
value-neutrality need be motivated by a desire to detach one’s analysis from
all ethical concerns. Value-independence might imply value-freeness and
value-neutrality, but the converse is certainly not the case.
Although I have no reason to take issue here with value-independence as
a methodological desideratum, I shall not discuss its validity. Indeed, I shall
assume its rejection, arguendo. My concern here is to defend value-freeness
and value-neutrality as features of analyses that make no pretense to being
wholly detached from ethical concerns. Even if value-independence is
impossible or undesirable, even if our analyses are inevitably expressions of
our ethical concerns, there are strong reasons for favoring value-freeness and
value-neutrality (though not always in combination) in the analysis of cer-
tain concepts in the light of certain theoretical aims. Value-freeness and
value-neutrality have important roles to play in ethical theorizing.
In section 1 of this chapter, I shall summarize some of the reasons that
have been provided for affirming or denying the possibility or usefulness of
value-freeness, value-neutrality, and value-independence—three features of
concepts that the methodological debate so far has tended to roll into one,
or at least to see as mutually entailed. Building on this summary, I shall then
define and distinguish the three features. In section 2, I shall examine and
defend the notion of value-freeness, as distinct from value-neutrality and
value-independence. In section 3, I shall examine and defend the notion of
value-neutrality, as distinct from value-freeness and value-independence. In
the process, I hope to show both how value-freeness and value-neutrality
matter, and how they can come apart.
I shall assume throughout that the term “value,” in expressions like “value-
free,” “value-neutral,” and “value-independent,” refers exclusively to ethical
value. In other contexts one might of course refer to non-ethical values
(including the scientific values of coherence and clarity), but in the present
context, in line with the relevant literature, I confine my attention to ethical
values. Likewise, I use the term “evaluative” to mean “ethically evaluative.”
1. VALUE-FREENESS, VALUE-NEUTRALITY, AND

VALUE-INDEPENDENCE
A number of philosophers have defended the view that political concepts fall
into two categories: the evaluative and the non-evaluative. On this view,
freedom and power are prime examples of non-evaluative political concepts.
282 Ian Carter
Other concepts that have sometimes been said to fall into this category
include equality, democracy, and law. Evaluative concepts, on the other
hand, include justice and desert: unlike in the case of freedom and power,
one cannot say that a certain distribution of resources is just, or that a certain
person deserves a reward, without thereby making a normative ethical claim.
Authors who have taken this dualist or “separationist” view of political con-
cepts have included Felix Oppenheim (1961, 1981), Hillel Steiner (1994),
Matthew Kramer (2003), and Frank Lovett (2010). According to Oppenheim,
for example, we should define social freedom in the following way: agent A
is free to perform some action x if and only if no other agent is rendering A’s
performance of x impossible or punishable (1981, p. 53). Whether or not
some other agent is rendering x impossible or punishable can be ascertained
empirically without recourse to value judgments (as long as we can in turn
plausibly understand the concept of punishability in purely empirical
terms1). Am I free to attend the church down my road, or to vote, or to
leave the country? The answer depends on whether some other agent is
making it impossible for me to do so, or would successfully punish me were
I to do so. Statistical evidence will provide us with empirical (probabilistic)
answers to such questions. Freedom is, therefore, a different kind of concept
from the concepts of justice and desert.
But this view has been hotly contested by those who see all political con-
cepts as necessarily value-laden. According to these authors, who include
William Connolly (1993), Steven Lukes (2005), and Ronald Dworkin
(1986, 2004, 2011), freedom, power, democracy, and law, no less than jus-
tice, desert, and the good life, must be analyzed and defined as part of a
system of ethical values. Each of them is inescapably an ethical concept.
Thus, in the case of freedom, these authors hold that one cannot specify
what a person is free or unfree to do without first specifying what it is valu-
able for him or her to do, and one cannot compare the degree of freedom of
one person with that of another without comparatively evaluating the ends
that those freedoms permit them to realize.
Those in the first camp have sometimes claimed that their non-evaluative
definitions of concepts like freedom or power are neutral between compet-
ing ethical standpoints. This “value-neutrality” is said to be useful because it
provides us with a shared starting point in terms of which to express genu-
ine ethical disagreements. Oppenheim illustrates this point with the follow-
ing example. Imagine that in my opinion freedom is best realized under
1
One person’s punishment might be another person’s reward. However, such evalua-
tions of sanctions can themselves be objects of empirical investigation. Steiner (1994),
Carter (1999), and Kramer (2003), on the other hand, define unfreedom only in terms
of physical impossibility.
classical liberalism whereas in your opinion freedom is best realized under

socialism. If we each mean different things by “freedom,” no meaningful
substantive disagreement is expressed by these apparently contrasting
claims. We are simply talking past each other, using the word “freedom” to
refer to different concepts. If, by contrast, we have already agreed on what
freedom is, but we disagree about which of these two political systems best
promotes freedom, or promotes the best kinds of freedom, then we are
disagreeing meaningfully (about the empirical conditions for the realization
of freedom, or about the value of particular freedoms or kinds of freedom).
Suppose now that a reliable empirical investigation reveals classical liberal-
ism to provide more freedom, or to better guarantee certain kinds of free-
dom, than socialism does (on our agreed definition of freedom). Such a
result, in Oppenheim’s view, does not in itself tell in any way against social-
ism. “On the basis of the proposed ‘value-neutral’ language, to deny that
poverty or unemployment under capitalism is an instance of social unfree-
dom is not to justify capitalism” (1981, p. 55–6; 1995, p. 416).
Those in the second camp deny that any political concept can be value-
neutral in this way. The point of political concepts like freedom, power, and
justice, they say, is to make ethical evaluations. Thus, even if one attempts
to define and use such concepts non-evaluatively, one will still be making
an ethical point, whether one likes it or not. Ethical neutrality in the analy-
sis of such concepts is therefore an illusion. Logically speaking, there is of
course nothing to stop the socialist, in the example given above, from
declaring herself “against freedom” and continuing to support the same pol-
icies at the material level. But to do so would be to concede a normative
point to the liberal. Unless, that is, the liberal also gives up her point of view
on the value of freedom. And were the socialist and the liberal both to give
up their points of view on the value of freedom, we would be left with a
rather different concept (of freedom) from the one we originally set out to
define (Connolly 1993, pp. 32–3). As Dworkin writes, “politicians who
defend taxation do not concede that taxation invades freedom.” If the defi-
nition is really neutral, “then why should politicians and citizens waste time
arguing about it?” (2004, p. 7).
Dworkin diagnoses the illusion of value-neutrality as the product of a
general view he labels “archimedeanism.” Archimedeanism is a metaethical
stance according to which we can “stand outside a whole body of belief
[in this case, ethical belief ], and . . . judge it as a whole from premises or
attitudes that owe nothing to it” (Dworkin 1996, p. 88). Dworkin’s primary
target is H. L. A. Hart’s concept of law (1994), but he applies the same
argument to all ethical and political concepts. According to Dworkin’s
anti-archimedeanism, such concepts cannot be correctly understood inde-
pendently of an interpretation of the ethical practices within which they are
284 Ian Carter
used. They cannot be correctly defined from some perspective wholly out-
side ethics. Ethical and political concepts, including freedom no less than
justice, are “interpretive” concepts, and are in this sense different from the
“criterial” concepts used in the empirical sciences (Dworkin 2004; 2011,
Ch. 8). The concept of a lion can be explicated by setting out empirical
criteria, such as the DNA of lions. The concepts of freedom and justice
cannot be explicated in this way, but are properly interpreted and clarified
by reference to the system of ethical values of which they are a part. For
example, if the ideal of freedom is part of an egalitarian ethical vision of
society (as it is for Dworkin), it needs to be interpreted and defined in terms
of egalitarian values.
I suggest that there are three distinct notions at work in the above debate,
and that these notions can come apart in interesting ways. First, there is the
notion that I call “value-freeness.” This notion can be defined as follows:
Value-freeness: a concept is value-free if its definition is such that the
definiens contains no evaluative terms.
I assume that an evaluative term expresses an evaluative concept. When moral
philosophers refer to an “evaluative concept” they usually mean a concept the
use of which necessarily involves an evaluation, or what I shall here call an
“essentially evaluative concept.” There are, however, many concepts that are
not themselves evaluative in this strict sense but are often used evaluatively.
These are concepts the use of which does not necessarily express an evaluation,
but often has evaluative connotations. I shall call these concepts “non-essen-
tially evaluative.” Hillary Putnam (1981, p. 209) provides the following exam-
ple: “in our culture ‘slobbers his food all over his shirt’ has strong negative
emotive force although the phrase is literally a description.” I shall assume here
that a definition counts as value-free if the definiens is free not merely of essen-
tially evaluative terms but also of non-essentially evaluative terms—that is, of
evaluative terms in the broader sense. More generally, when I refer to “evalu-
ative concepts” without further qualification, I should be taken to refer to
evaluative concepts in this broader sense. In addition, I assume that what
counts as a non-essentially evaluative concept (as opposed to a straight for-
wardly non-evaluative concept) depends on the evaluative stance of the per-
son using it—that is, it depends on whether the concept serves, from its user’s
ethical point of view, to provide evaluations of the world.
When Oppenheim claims that freedom and power should be given
“empirical” definitions, or are “purely descriptive” concepts, he means that
they should be treated as value-free in my sense. Similarly, when Connolly
and Dworkin claim that these concepts are evaluative, they mean that they
are essentially evaluative in my sense and cannot or should not be treated as
value-free in my sense.
Second, there is the notion of value-neutrality, which I define as follows:
Value-neutrality: a concept is value-neutral if its use does not imply the

superiority of any one of a set of contrasting substantive ethical points
of view.
Unlike value-freeness, value-neutrality is a matter of degree, for value-
neutrality always exists relative to a particular range of substantive ethical
points of view, and this range can vary in breadth. Moreover, since val-
ue-neutrality pertains to different ethical points of view, it is always relative
to a given context of shared meanings in which various members of a set of
concepts are understood to have various evaluative implications.
Since I shall assume, arguendo, the value-dependence of all ethical and
political concepts (an issue discussed immediately below), I shall assume that
it is not possible to define an ethical or political concept in a way that is abso-
lutely value-neutral. While many ethical and political concepts display at least
some degree of value-neutrality when employed in certain contexts, I shall
assume that no such concept can be neutral with respect the set of all conceiv-
able ethical points of view. This is not to deny, on the other hand, that the
relevant range of ethical points of view might be extremely broad—say, for
example, all actually affirmed contemporary Western views about the nature
of just institutions or of a just distribution of resources. As we shall see in
section 3, one might try to achieve this degree of value-neutrality by adopting
an extremely abstract concept of justice. I shall call this kind of value-neutral-
ity “metatheoretical value-neutrality.” A more limited, but still important,
kind of value-neutrality can be achieved within normative theorizing itself—
as, for example, where political liberals embrace a definition of justice that is
neutral between a wide range of views about the nature of the good life. I shall
call this kind of value-neutrality “normative value-neutrality.”
When Oppenheim defends empirical definitions of freedom or power by
claiming that they allow for a consensus on the basis of which to understand
normative political disagreements, his argument is premised on the useful-
ness of value-neutrality in my sense of the term. Similarly, when Connolly
and Dworkin deny that a definition of freedom like that provided by
Oppenheim can be acceptable from such different ethical points of view as
liberalism and socialism, they are denying the possibility of value-neutrality
in my sense of the term. Both sides in this debate seem to have been assuming
only an absolute notion of value-neutrality.
Third, there is the notion that I call “value-independence,” and which
I define as follows:
Value-independence: a concept is value-independent if its definition can
be justified purely in terms of theoretical-explanatory considerations,
and not at all in terms of ethical considerations.2
2
I take the term “value-independence”, and its meaning, from Kramer (forthcoming).
286 Ian Carter
To say that a given political concept is value-independent is to say that,

although it might take on evaluative meanings in everyday political dis-
course, it can nevertheless be defined simply with a view to improving our
understanding of empirical phenomena and without presupposing or refer-
ring to any of those evaluative meanings.
When Dworkin attaches the label “archimedeanism” to the attempt to
construct value-free or value-neutral political concepts, he really has in
mind what I am here calling value-independence.3 Oppenheim’s account of
freedom would certainly qualify as archimedean in Dworkin’s eyes. Under
the influence of logical positivism, Oppenheim aimed to appropriate the
concept of freedom as a tool for social scientists, defining it in a way that is
motivated not by reference to any ethical considerations but by reference to
the criteria for the fruitful pursuit of empirical knowledge. Although he
thought that his definition would be useful in clarifying the differences
between normative theories (a point illustrated earlier), his aim in defining
freedom empirically was not to defend or support or reconstruct any kind
of normative theory. Instead, he hoped that social scientists might use his
analysis of freedom to establish empirical correlations between freedom and
other social and political phenomena, such as democracy or stability or eco-
nomic development, and thereby improve our understanding of political
events and states of affairs.
Value-independence is easily confused with value-freeness and value-
neutrality, but the three features are analytically distinct. To affirm the
value-independence of an ethical or political concept is to make a point
about the justificatory grounds (ethical versus non-ethical) on which to pre-
fer one definition of that concept over another. To affirm its value-freeness
is to make a point not about the role of values in justifying a definition, but
about the presence or absence of values in the definition itself. To affirm its
value-neutrality is to claim that it can be used in given contexts without
implying the speaker’s allegiance to one or another member of a set of sub-
stantive ethical points of view.4
In light of these differences, it is also worth noting how the three features
differ in terms of whether they are intrinsic or extrinsic features of a con-
cept. Value-freeness is an intrinsic feature of a concept, depending as it does
on the nature of the definition of the concept. Value-independence and
3
Dworkin’s notions of “austerity” and “neutrality” (1996) bear some resemblance,
respectively, to my notions of “value-freeness” and “value-neutrality.” However, for
Dworkin, austerity and neutrality are themselves metaethical notions that characterize
the position he calls “archimedeanism,” and they come apart only in the case of moral
error theory (which affirms austerity without neutrality).
4
Value-independence seems to be the closest of my three notions to Weberian Wertfreiheit,
even though the latter is normally translated as “value freedom” or “value neutrality.”
value-neutrality, on the other hand, are extrinsic features of a concept,

depending, respectively, on the justification of its definition and on its
implications in given contexts.
2. VALUE-FREENESS
In this section, I provide a defense of the notion of value-freeness that does

not assume value-independence. I agree with Oppenheim that we should
distinguish between concepts like freedom and concepts like justice in
terms of the value-freeness of the former, but I think that we can make this
distinction from within a particular ethical perspective. Once we see more
precisely how value-free concepts are distinguished from non-value-free
concepts within a particular ethical perspective, we shall also see the impor-
tant role that the former can and should play in ethical theorizing.
I shall first argue against the view that the distinction between concepts
like freedom and concepts like justice corresponds to that between descrip-
tive concepts and evaluative concepts. The concept of freedom is both
descriptive and evaluative, and the concept of justice is both evaluative and
descriptive (section 2.1). Instead, I shall suggest, the difference lies in the
relative locations of these evaluative properties in a hierarchy.5 The kind of
hierarchy I have in mind is one that expresses a grounding relation (sec-
tion 2.2). I shall illustrate this grounding relation by reference to the par-
ticular ethical outlook of a liberal who counts freedom and justice as among
her political values. As we shall see, the relations between the different eval-
uative properties located in the hierarchy tell us about the sorts of reasons
the liberal gives for saying that qualities like freedom or justice are present
in the world, or would be present if we were to change it in certain ways
(section 2.3). Value-free concepts play an important role in such normative
reason-giving, because they provide a bridge between our value systems and
the empirical world (section 2.4).
2.1 Description and Evaluation

Concepts that I am here calling “value-free” have sometimes been charac-
terized as “descriptive.” However, my primary concern in the present con-
5
I use the term “properties,” here, simply to mean those characteristics that we must
recognize objects as having in order to refer to the ways in which they are alike or differ
from one another. This “superficial” usage of the term “property” neither presupposes nor
contradicts realism about ethical properties. A similar point applies where I speak of
freedom or justice as “present in the world.”
288 Ian Carter
text is with value-free concepts that are themselves evaluative in the broad
sense. Moreover, even essentially evaluative concepts are used to describe.
The distinction between descriptive and evaluative concepts is therefore
potentially misleading.
This is not to deny that there is an important distinction between descrip-
tion and evaluation. Peter Morriss has illustrated this distinction by consid-
ering the way in which we use the term “persecution.” It would be strange
to say: “A lot of persecuting is going on and all the better for it.” In making
a statement about the presence of persecution in a political situation, I am
automatically taken to be presenting a (prima facie) negative evaluation of
that situation. However, Morriss goes on, if the activities of evaluation and
description were not distinct, I could not be taken to have a reason for my
evaluation. When I say that a particular society contains persecution, what
I am saying is in fact short for: “I condemn this situation (evaluation) and
the reason I condemn it is that it involves persecution (description)”: “The
speaker has to have some descriptive content in mind when using the term
‘persecution’, otherwise he is just saying that he condemns the society for no
reason at all” (Morriss 2002, p. 201).6
Morriss is certainly right to distinguish in this way between the evaluative
and descriptive content of evaluative judgments. However, the distinction
will not suffice to further the cause of value-freeness in the analysis of polit-
ical concepts, for the concepts of freedom and justice each have both a
descriptive and an evaluative function: if a liberal describes a country as
“free,” she implicitly (and prima facie) evaluates the country in a positive
way; if a champion of justice evaluates the outcome of a criminal trial as
“just,” she describes that outcome as, say, something that has come about as
the result of a certain procedure having been followed (or, if the outcome is
considered in isolation from the procedure that produced it, as a result that
reflects the truth about a person’s guilt or innocence). Claims about free-
dom both describe and evaluate; and claims about justice both describe and
evaluate. There is nothing less material about justice than there is about
freedom. It is true that when we ascribe freedom or power to a person or
situation, we are saying something definite about the non-evaluative, empir-
ical world: we are saying that certain material pushings, pullings, influences,
6
Moral philosophers echo this point when they say that a “thick” ethical concept (of
which persecution is presumably an example) can be analyzed into a “thin” evaluative
component and a descriptive component. I shall not make use of the distinction between
“thick” and “thin” evaluative concepts here, as I share Samuel Sheffler’s misgivings about
the clarity of such a dichotomy. Is justice a “thick” or a “thin” evaluative concept?
(Sheffler 1987, p. 417). Moreover, the hierarchical model I present below (in section 2.2)
implies that all evaluative concepts, including supposedly “thin” ones, can have descrip-
tive content.
psychological conditionings, or whatever, have taken place or are taking

place or might take place. But the same is true of an evaluative concept like
justice: if we say that there is something unjust in the situation of Jones, we
must mean, for example, that Smith is doing something unjust to Jones,
and the “doing” in question must, ultimately, be describable in terms of the
occurrence of certain empirical (physical or mental) events.
2.2 Grounding Hierarchies

The “evaluative-descriptive” dichotomy seems to be unhelpful when trying
to make sense of the Oppenheimian intuition that there is a relevant differ-
ence between concepts like justice and concepts like freedom. Nevertheless,
the foregoing reflections do point towards an alternative formulation of
that difference. To arrive at this alternative formulation, we need to think
of our concepts as arranged in a hierarchy ascending from the more specific
to the more general (Kovesi 1967). The presence of the properties picked
out by the more specific concepts serves to ground the presence of the proper-
ties picked out by the more general concepts. The grounded properties are
present in virtue of the presence of the grounding properties, and not vice
versa. For example, it might be said that a person qualifies as good in virtue
of her qualifying as generous, and that she qualifies as generous in virtue of
her qualifying as having made certain gifts to certain people. The more
general properties located near the top of the hierarchy will often be
grounded in conjunctions and disjunctions of properties that are located
lower down, making the properties picked out by the higher-level concepts
more complex than those picked out by the lower-level ones. The complete
set of properties referred to by a particular ethical outlook will amount to
a grounding network.
This ontological grounding relation has often been characterized as one
of supervenience—in particular, of “strong,” or inter-world supervenience
(Stalnaker 1996; McLaughlin and Bennett 2011). However, the superven-
ience relation does not possess all of the properties possessed by the ground-
ing relation that I have in mind: while both relations are transitive, only the
grounding relation is asymmetric and irreflexive (Correia and Schnieder
2012). I shall therefore continue to characterize the relevant hierarchy or
network in terms of grounding rather than supervenience.
In the process of constructing a substantive normative theory, we aim to
clarify the nature of an evaluative property by looking downwards or
upwards or sideways in the hierarchical network in which we suppose it to
be located. When we look downwards in the hierarchy, we point to exam-
ples that suggest what sorts of things fall under a given concept. When we
look upwards in the hierarchy, we point to the wider categories of which we
290 Ian Carter
believe a concept to be a part. When we look sideways at other concepts

located at the same level, we point to sets of concepts that seem to share
those same wider categories. When we combine these ways of clarifying a
concept we engage in a kind of holistic normative analysis—a species of
what P. F. Strawson calls “connective analysis” (Strawson 1992, Ch. 2).
Engaging in this kind of analysis involves attempting to shed light on the
nature of each property in terms of its position with respect to other prop-
erties in the grounding network. In this way, we clarify the point of concepts
like those of freedom or justice. We gain a clearer understanding of their
evaluative connotations, some of which were brought out in the discussions
about freedom cited in the previous section.
2.3 A Liberal Grounding Hierarchy

Consider now a broadly liberal ethical standpoint. (I mean here a stand-
point that is liberal in an ideologically broad sense, and not only the “clas-
sical” liberal standpoint referred to in the example given in section 1.) From
this standpoint it is plausible to think of justice as grounded in (among
other things) there being a certain amount or distribution of freedom. In
this sense, freedom grounds justice, and not vice versa. Thus, from a liberal
normative political point of view, freedom and justice occupy a single hier-
archy of grounding and grounded ethical properties. But the places they
occupy are very different. I suggest, indeed, that when political liberals
theorize carefully and systematically about the proper relations between
the terms they use to articulate their normative political ideals, their recon-
struction of the relevant hierarchical network ought not only to place jus-
tice at a very high level, but also to place freedom squarely at the bottom.
As I shall now try to show, on the basis of this suggestion a liberal concept
of freedom can be plausibly understood as both value-dependent and value-
free—that is, as an integral part of a liberal system of values, as defined
from the perspective of that system, and yet as defined in terms that are not
themselves evaluative.
The particular portion of our hierarchical network that has justice at its
summit will represent a particular conception of justice. The difference between
defining a concept and pointing to its grounding properties indeed provides
a useful clarification of the well-known Rawlsian distinction between a
“concept” and its various “conceptions.” Defining the concept of justice
might involve saying, for example, that justice consists in giving each person
her due.7 Different “conceptions” of justice will then point to the properties
7
This is not, of course, “the concept of justice” as Rawls himself defines it at p. 5 of
A Theory of Justice (1999). I return to Rawls’s own definition in section 3.3.
that ground those referred to in the definiens—in this example, the property
of “being due to a person.” The grounding properties tell us what counts as an
instantiation of the grounded property. The political liberal might say, for
example, that giving each person her due consists in respecting her rights:
respect for rights is what is due to people, it is what counts as people’s due;
justice is grounded in (that is, it exists in virtue of ) respect for people’s rights.
She might then go on to ground rights in the satisfaction of certain funda-
mental interests. And she might then continue in her explication of her con-
ception of justice by arguing that one such fundamental interest consists in
an interest in freedom. This substantive conception of justice would be
advocated by appeal to a so-called “interest theory” of rights together with
a justification of the fundamental interest in freedom: justice is grounded in
respect for rights, is grounded in the satisfaction of interests, is grounded
in the protection or promotion of freedom.
Elsewhere I have defended this ethical construction by appeal to the
“non-specific” or “content-independent” value of freedom (Carter 1995,
1999; cf. van Hees 2000; Kramer 2003). The non-specific (or content-inde-
pendent) value of freedom is the value possessed by freedom independently
of the fact of its consisting in the freedom to do one or another specific
thing. It is the value that freedom has as such. If people have an interest in
freedom as such, there will be contexts in which it is appropriate to increase
or promote their freedom. Their rights protect, among other things, their
interest in having a measure of freedom. And this measure must be ascer-
tained empirically, without reference to the value of the freedom specifically
to do this or that thing (Carter 1999, Part 3).8 This particular theory of
freedom provides just one example of the way in which one might justify
locating a particular property at the bottom of one’s grounding hierarchy of
evaluative properties.
Some evaluative properties, then, are located at the bottom of a grounding
hierarchy of evaluative properties, and are not themselves grounded in eval-
uative properties. Consider our example of a liberal theory of justice as
grounded in a right to freedom. If we continue to descend beyond this hier-
archy of evaluative properties, we shall pass from particular instantiations of
8
Because Dworkin ignores the reasons we have for valuing freedom non-specifically,
he cannot see how freedom could be located at the bottom of the hierarchy without los-
ing its ethical value. And yet he cannot avoid locating freedom, implicitly, at the bottom
of the hierarchy. He therefore distinguishes between mere empirical freedom and the
truly ethical value of “liberty,” which he implicitly locates much higher up in the hierar-
chy (Dworkin 2011, Ch. 17). It is not clear what this moralized concept of liberty adds to
an interpretation of justice based on equality, rights, and empirical freedom. Neither is it
clear how much conceptual significance should be assigned to a linguistic distinction that
can only be made in English.
292 Ian Carter
freedom to the properties of a set of particular bodily movements and other

physical and/or mental events the presence of which itself grounds those
instantiations of freedom, and then on down eventually to a set of micro-
physical properties that ground those properties of bodily movements and
other events. These empirical properties do not, of course, lack ethical signif-
icance: the bodily movements and other empirical phenomena in which free-
dom is grounded do have ethical significance insofar as their presence
grounds the presence of a property that we refer to in making evaluations.
However, the empirical properties are not themselves picked out by concepts
that are used evaluatively. To suggest that ethically significant states of affairs
are ethical “all the way down” is not to object to this last point. Where we
draw the line between concepts that are used evaluatively and concepts that
are not, will depend on the particular substantive ethical theory in question.
In the case of the political liberal referred to above, it will depend on her
particular conception of justice, which itself includes all the concepts she
uses in characterizing the nature of justice on that conception.
Moral philosophers are in broad agreement that all ethical properties are
ultimately grounded in empirical properties (Hare 1984; Jackson 1998;
Smith 2000; Kramer 2009).9 What matters for present purposes is that
some such ethical properties are grounded in empirical properties directly,
whereas others are grounded in them only indirectly. As Julius Kovesi
observes in his illuminating discussion of conceptual hierarchies, “[i]t seems
obvious that the higher a term is [in a hierarchical structure] . . . the more
difficult it is to specify the empirically observable qualities, aspects and rela-
tions or movements of things or bodies that have to be present for the
proper use of that term” (1967, p. 34). This observation explains why we
need to refer to the mid-level and lower-level properties in our grounding
hierarchy. Although the grounding relation is transitive, and although it
therefore remains true that a property that is located high up in a hierarchy
of ethical values is ultimately grounded in a set of empirical properties (a set
of “empirically observable qualities, aspects and relations or movements of
things or bodies”), it is generally impossible for us, in practice, to verify the
presence of that lofty ethical property by pointing directly to those empirical
properties. The relation “is grounded in” is transitive, but the relation “is
identifiable in terms of the presence of ” is not similarly transitive. The most
general and complex ethical properties are therefore identified empirically
by pointing to the presence of intermediate ethical properties in which they
9
According to Kramer (2009, p. 211), the thesis that moral properties are “generated
by” (or strongly supervene on) empirical properties is “unexceptional,” but is also “a
profoundly ethical thesis.” If so, that thesis can be understood unproblematically as part
of the overall ethical point of view on which our value-free definition of freedom depends.
are grounded in a more immediate sense, where these more immediately

grounding properties are in turn grounded in other, yet more specific ethi-
cal properties.
The grounding properties in the hierarchy serve to indicate what Morriss,
as cited above, referred to as the “descriptive content” of ethical claims, or
what Kovesi calls the “recognitors” of the property in question—that is, “the
defining characteristics of the material elements of a thing,” where the pres-
ence of those material elements is what shows the thing itself to be present
(Kovesi 1967, p. 41). One can imagine, for example, the following conversa-
tion in which a series of such reasons emerges as we descend the liberal
grounding hierarchy step by step: Q: “Why did you describe that state of
affairs as a just one?”; A: “Because those people’s rights have been respected”;
Q: “Why do you describe the state of affairs as one in which those people’s
rights have been respected?”; A: “Because these actions of these people consti-
tute a protection of the fundamental interests of those people” . . . and so on.
At each step, an immediately grounding property is appealed to in justifica-
tion of a description made in terms of an immediately grounded property.
If we follow this chain of reasons in support of our descriptions of the
world, we shall eventually arrive at an evaluative claim the descriptive con-
tent of which is appropriately specified exclusively in non-evaluative terms.
The property referred to in such a claim will not itself be an essentially eval-
uative property, but it will nevertheless constitute the last evaluative element
in an asymmetric chain of descriptive reason-giving. In the example I have
used to illustrate the grounding hierarchy, that of political liberalism on my
own favored interpretation, freedom qualifies as just such a property: Q:
“Why do you describe this person as having such a large extent of freedom?”;
A: “Because, while these obstacles prevent her from doing a, b, and c, she
nevertheless remains unprevented from doing x, y, and z; s, t, and u; . . . .”
From within this liberal perspective one might say, with Bernard Williams,
that freedom is a “primitive” term—not, of course, in the sense that it is
unanalyzable, but in the sense that it refers to a “quite basic human phe-
nomenon” (Williams 2001, p. 11; cf. Ricciardi 2007, p. 137). In other words,
the concept of freedom is closer to those of biology and physics than are
descriptively less basic concepts like justice.
From within this particular liberal perspective, the difference between
freedom and justice is not that the former concept is value-independent
whereas the latter is value-dependent. Neither is it that the former is value-
neutral whereas the latter is not (as we shall see in section 3, value-neutrality
is a feature that might or might not be ascribable either to freedom or to
justice, depending on how, and with what theoretical aims, these concepts
are defined and used). Rather, the difference is that the concept of freedom
is best understood as picking out empirical features of the world directly,
294 Ian Carter
whereas the concept of justice is best understood as picking out evaluative

features directly, and empirical features only indirectly. Both concepts are
ethical. And yet the difference in question makes it plausible to place free-
dom, and not justice, within the category of value-free concepts. Freedom
is a distribuendum of justice, because freedom is a property that has
non-specific value. Only a value-free concept of freedom will pick out free-
dom understood as that kind of property.10 This is not to say exactly what
the best definition of freedom is, but it does narrow down the set of candi-
date definitions. Those of the right kind will include the definitions set out
by Oppenheim (1995) and Steiner (1994). From the liberal point of view
just described, such definitions serve to identify concepts that are value-free,
but not value-independent.11
2.4 Why We Need Value-free Concepts

The view of freedom described above happens to be one that I endorse, but
I have set it out primarily by way of illustration. The main point I wish to
make is more general, and concerns the need for value-free concepts in
normative political theory.
Value-free concepts provide a bridge between our value systems and our
ways of carving up the material world, so that our claims about what is valu-
able, and what ought to be done, can make a difference to how we see and act
in that world. If there were no such bridge between our value systems and the
empirical world, those value systems would have no practical relevance. If all
of our values were defined in terms of other values, in a circular fashion, rather
than in terms of empirical categories, they would float free of the empirical
world and would fail to tell us what to do. They would be connected to each
other, but they would fail to “touch down in the realm of actual experience”
(Lovett 2010, pp. 18–19). Hence, the insistence among moral philosophers
that the evaluative supervenes upon, or is grounded in, the empirical.
10
Matthew Kramer (forthcoming) summarizes this position effectively, although he
uses the term “value-neutral” where I am here using the term “value-free” (I too have used
the term “value-neutral” in this way in some previous writings).
11
In his critique of value-free definitions of freedom (or what he calls “opportunity”),
Robert Sugden objects that there is no “neutral perspective” in terms of which we can
measure people’s available action, and concludes, skeptically and paradoxically, that
although we value opportunity as such (or as I would put it, although we ascribe non-
specific value to freedom), it is impossible to say how much opportunity (or freedom)
there is in the world (Sugden 2003, p. 802). Presumably, by a “neutral perspective,”
here, Sugden means what I would call a “value-independent” perspective, in which case
the distinction between value-freeness and value-independence ought to save us from
his puzzling conclusion. This is not to deny that the empirical measure I proposed (in
Carter 1999) might require revision.
It might be objected that we can dispense with value-free concepts with-

out allowing the set of evaluative properties to float free of the empirical
world. For it might be claimed that evaluative properties are grounded col-
lectively in empirical ones. On this view, although each and every ethical
property is defined in terms of other ethical properties, the complete set of
such interdefined ethical properties is nevertheless grounded in a certain
configuration of the empirical world. But it is not clear that such a collective
grounding relation can give our values practical relevance, for our ability to
perceive the material implications of any evaluative judgment will then
seem to depend on monumental powers of reflection covering the entire
expanse of a coherent system of values. This is not a plausible account of
how our ethical concepts acquire practical relevance.
One might endorse a more moderate view of the interdefined nature of
ethical and political values: a pluralist view, according to which various
smaller, more manageable clusters of interdefined values are each separately
and directly grounded in a set of empirical properties, so that there is again
no single value-free concept in my sense. At this point, however, one may
legitimately ask for a principled distinction between a single property and a
cluster of evaluative properties. If one can immediately grasp the way in
which a particular cluster of properties is directly grounded in a particular
set of empirical properties without the help of more specific intervening
evaluative properties located lower down in the hierarchy, then one ought
to be able to define that cluster in a value-free way.
Dworkin favors interdefined ethical properties and he rejects pluralism,
but he seems to avoid the counterintuitive consequences of a collective
grounding relation, mentioned above. I believe that he does this by implic-
itly using value-free concepts to anchor his overall value-system to the
empirical world. He rejects empirical definitions of liberty, preferring to
define liberty as the possibility of doing what one is able to do in an ideally
egalitarian society. But he grounds an ideally egalitarian society in individ-
uals’ equal possession of resources, and he measures resources in terms of
their market values. Once this grounding hierarchy has been made more
explicit, it is reasonable to expect that somewhere along the line we shall
find a concept that is defined empirically but used evaluatively.
Most of Dworkin’s methodological prescriptions simply fail to contradict
the bridging role that I have assigned to value-free concepts. Dworkin him-
self seems to think that the rejection of value-free concepts follows from the
unavailability of an archimedean perspective on ethical and political concepts.
But to reject the archimedian perspective is to reject the value-independence
of ethical and political concepts, not the possibility that some of them are
value-free in my sense. Dworkin also appeals to the idea that ethical and
political concepts must be interpreted holistically: “If we are better to
296 Ian Carter
understand the non-instrumental integrated values of ethics,” he writes, “we

must try to understand them holistically and interpretively, each in the light
of the others, organized not in hierarchy but in the fashion of a geodesic
dome” (Dworkin 2004, p. 17). We have already seen, however, that value-
free concepts can be clarified and defined by means of a holistic or “connec-
tive” approach to normative analysis.
As a matter of fact, geodesic domes have tops and bottoms, and therefore
provide a nice metaphor for what I have in mind in talking of a grounding
hierarchy with a bridge to the empirical world. Better still would be the
metaphor of a traditional, non-geodesic dome, like Brunelleschi’s dome in
Florence, Italy. A maximally general evaluative property (presumably, that
of goodness) is like the pinnacle of such a dome, from which other stones
(other evaluative properties) descend in various directions. A more specific
and non-essentially evaluative property (such as freedom or power) is like one
of the stones that together form the base of such a dome, and the empirical
properties grounding that more specific property are like the lower stones,
and ultimately the land, upon which the dome in turn rests.
3. VALUE-NEUTRALITY
I have defined a value-neutral concept as a concept the use of which does

not imply the superiority of any one of a set of contrasting ethical points of
view. I have conceded to Connolly and Dworkin, arguendo, that such value-
neutrality cannot exist in the absolute sense in which Oppenheim sees his
concepts of freedom and power as value-neutral. Absolute value-neutrality
presupposes value-independence. Or so I am assuming. Nevertheless, some-
thing of the spirit of Oppenheim’s approach can be preserved in the form of a
relative notion of value-neutrality, according to which value-neutrality
exists in degrees that vary from one value-dependent concept to another. As
we shall see, certain value-dependent concepts can exhibit a very high
degree of value-neutrality.
Ethical and political concepts with a high degree of value-neutrality have an
important role to play in both our normative and metatheoretical reflections.
I shall focus on these two areas of reflection, respectively, in sections 3.1
and 3.2. In both cases, the method by which we identify value-neutral
concepts is the same: it consists in revealing, or constructing, areas of inter-
section between contrasting grounding networks. In the normative case,
we seek intersections by aiming to identify shared normative concepts
within otherwise divergent value-systems. This involves an extension of
Oppenheim’s strategy from the empirical level to the normative level. In
the metatheoretical case, we seek intersections by abstracting from the dif-
ferent empirical or normative concepts used in divergent value-systems.

The resultant abstract properties are themselves grounded in a disjunction
of less abstract properties recognized within those divergent systems. As
we shall see, a high degree of value-neutrality can be a feature either of a
value-free concept (or of an abstraction from different value-free con-
cepts) or of a non-value-free concept (or of an abstraction from different
non-value-free concepts).
What is the point of value-neutrality? In the normative case, the point is
to allow disagreeing parties to converge on a particular ethical evaluation or
prescription. This convergence can itself be a normative desideratum, espe-
cially (but not only) for political liberals. In the metatheoretical case, the
point of value-neutrality is to reveal a basic conceptual structure that is
common to a set of different concepts. The discovery or construction of this
common conceptual structure can be important both in clarifying the dif-
ferences between conflicting normative theories and, relatedly, in identify-
ing the abstract ethical categories that they share.
In a final section (3.3), I shall try to clarify the relation between metathe-
oretical value-neutrality and the Rawlsian distinction, already touched on
in section 2.3, between a “concept” and its various “conceptions.”
3.1 Normative Value-neutrality

When Oppenheim envisaged the parties to an ethical disagreement agree-
ing on an empirical definition of freedom or power, what he was imagining
was, in effect, an area of intersection between their otherwise divergent
grounding networks. The socialist and the liberal, in the example discussed
earlier, espouse different comprehensive sets of ethical values. Translated
into the language of grounding, Oppenheim’s idea is that these different sets
of values are grounded in empirical states of affairs which the parties would
do well to describe by means of shared empirical concepts like freedom and
power. For one person, a high degree of empirical freedom, or a certain kind
of empirical freedom, might ground justice or goodness; for another, it might
ground injustice or badness. The idea is that we can clarify this ethical disa-
greement by inviting the two people to embrace grounding networks that
overlap at the empirical level even though those networks diverge as they
ascend into the realm of ethical properties.
Oppenheim was attempting to employ the concepts of freedom and
power at a level in the hierarchy that was sufficiently low as to make them
maximally value-neutral (or, as he saw it, absolutely value-neutral). This
endeavor was criticized by many as amounting to an improper use of those
particular concepts. Even these critics ought to admit, however, that his
basic strategy in pursuing value-neutrality was correct insofar as it implic-
298 Ian Carter
itly involved aiming to establish an area of intersection between different

grounding networks. This strategy can be extended to the realm of ethical
concepts, even though Oppenheim himself did not contemplate doing
so. In this case, we shall be inviting disagreeing parties to embrace ground-
ing networks that overlap at higher levels. Whether we are working at the
empirical or at the evaluative level, our strategy will involve abstracting
an intersecting area from its place in each of the otherwise divergent
grounding networks in which it occurs. Moreover, whether we are work-
ing at the empirical or at the evaluative level, the kind of neutrality
achieved will be value-neutrality as I have defined it, albeit value-neutrality
of varying degrees.
For example, when it is said that the presence of justice is grounded in
the presence of an adequate measure of freedom for all, one thing the
speaker might have in mind is that such a conception of justice is compati-
ble with, and acceptable to, the different ethical visions of all the members
a set of parties to a given dispute about the nature of the good life. As
Rawlsians would put it, the justice of equal liberty (defined in a particular
way) can be affirmed from within the “comprehensive doctrine” of each
party. The relevant concept of freedom can therefore be employed by each
of the parties to the dispute about the nature of the good life. It can be
employed by each of them in the same evaluative or prescriptive way, while
leaving open which of the relevant comprehensive doctrines commands
their allegiance. In this way, the concept of freedom can be understood both
as value-dependent and as having a high degree of normative value-neutral-
ity—that is, as being employable in the formulation of a single set of evalu-
ations or prescriptions affirmed by all the parties to an ethical disagreement
(or set of ethical disagreements).
Normative value-neutrality is a feature of those concepts that occupy the
intersecting area common to each and every one of the comprehensive
grounding networks that together constitute the relevant domain of neu-
trality. These (more or less) value-neutral concepts need not necessarily be
value-free. A particular concept of rights can be neutral between various
contrasting normative theories, even though it is defined in evaluative terms
within each such theory.
3.2 Metatheoretical Value-Neutrality

Intersections between different grounding networks need not be sought
only for normative reasons. We might also point to such intersections
in order to shed light on the nature of human values. In the latter case,
we increase the value-neutrality of our concepts by making them more
abstract.
Consider, for example, the dispute about freedom between Isaiah Berlin
(2002) and Gerald MacCallum (1967). According to Berlin, there are at
least two “concepts of freedom”: the negative concept that characterizes the
liberal tradition (roughly speaking, freedom as the absence of humanly
imposed constraints) and the positive concept that characterizes other tra-
ditions such as socialism or communitarianism (freedom as self-mastery or
self-realization). According to MacCallum, on the other hand, here is only
one “concept of freedom”: freedom as a triadic relation between an agent,
constraints, and doings or becomings. Whether one is talking about nega-
tive or positive freedom may depend, according to MacCallum, on whether
the relevant agent is an individual (as in the liberal tradition), or a collectiv-
ity; it may depend on whether one characterizes the relevant “constraints”
more narrowly, so as to include only physical prevention or the use of threats
(as in the liberal tradition), or more broadly, so as to include various kinds
of social pressure, fear, ignorance, or uncontrolled passions; and it may
depend on whether the relevant doings or becomings cover the class of all
possible outcomes (as in the liberal tradition), or only those that the
“authentic” or “rational” agent would choose to realize. People refer to dif-
ferent kinds of “agent,” “constraints,” and “doings/becomings” when talking
about freedom. But, MacCallum says, ultimately they are talking about the
same thing: freedom, understood as a triadic relation.
Since Berlin and MacCallum are theorizing at different levels of abstrac-
tion, they are not really in disagreement but are arguing to different pur-
poses (Ricciardi 2007, pp. 129–30). Berlin’s purpose is to elucidate certain
influential and contrasting ethical visions that make use of different con-
cepts of freedom and to help us decide which of these visions is the most
convincing. His purpose is that of normative theorizing (albeit with a his-
torical dimension). MacCallum’s purpose is metatheoretical: it is not to take
sides in the dispute about whether negative or positive freedom matters
most, but to explain why it is that we call both of them concepts of freedom.
The idea is to point to a basic conceptual structure that effectively captures
and clarifies our shared sense that the advocates of negative and positive
freedom, despite using different concepts, are nevertheless all talking about
“the same thing.” Advocates of positive and negative freedom admittedly
use the word “freedom” to refer to quite different concepts; these different
concepts pick out different properties; and yet, grounded in the disjunction
of these properties is a single abstract property—the property that is picked
out by MacCallum’s concept of freedom. By jointly recognizing the signifi-
cance of this abstract property, the two sides come to see that they have
more in common than a mere use of the same word.
The kind of value-neutrality achieved by MacCallum is quite different
from that envisaged by Oppenheim. Oppenheim fills in the three abstract
300 Ian Carter
elements of MacCallum’s triadic formula with a particular content that will

allow the concept to be used in (non-vague) descriptions of the world, and
also makes the (widely contested) claim that this particular proposed con-
tent is value-neutral (in an absolute sense). MacCallum instead arrives at a
neutral account of “what freedom is” precisely by abstracting from such par-
ticular contents as they appear in contrasting definitions. He does not aim
to describe the world in terms of freedom.
As in the case of normative value-neutrality, metatheoretical value-neutrality
does not require value-freeness. Just as Oppenheim’s method of seeking
value-neutrality can be extended from the empirical to the normative level
(once we give up the requirement that value-neutrality be absolute), so
MacCallum’s method can be applied to families of non-value-free concepts
as well as to families of value-free ones (although it might run into difficul-
ties where the concepts are all located very high up in their respective
grounding networks, as we shall see in the next subsection). MacCallum’s
method can indeed be applied to families of concepts of which some mem-
bers are value-free and others are not, and his own concept of freedom is
indeed an example of such a hybrid case. Steiner (1994) defines liberty as
the absence of physical prevention. His definition of liberty is value-free.
Dworkin (2011) defines liberty as the possibility of doing what one would
be able to do in an ideally egalitarian society. His definition of liberty is not
value-free. Dworkin thinks that liberty should be defined in terms of the
other values within his own theory of justice, so that liberty is shown to be
preserved or promoted by a just, egalitarian distribution of property. Steiner
holds that such a non-value-free definition excludes the possibility of citing
liberty as a reason for preferring one distribution of property over another.
MacCallum’s value- neutral concept abstracts from this dispute about
whether or not liberty should be defined in a value-free way. It is neutral
between value-free and non-value-free definitions.
Why should we care about such abstract, neutral concepts? Why should
we think of MacCallum’s project as an important part of political philoso-
phy? The answer lies in the role it can play in the more general philosophical
enterprise of elucidating the nature of human values. The exercise of clarify-
ing or defining a normatively value-neutral concept is part of the wider phil-
osophical exercise of constructing a normative ethical theory—in this case,
a theory that comprises value-neutrality as itself an ethical value—with
direct practical implications. By contrast, the point of a metatheoretically
value-neutral concept is to abstract from differences between such norma-
tive ethical theories and to say something more general about the kinds of
ethical concepts people use. At its most ambitious, it might even be to say
something interesting about “the human condition.” In this case, an abstract,
metatheoretically value-neutral concept will correspond to what Berlin
would have called a “category.” For Berlin, “there exist central features of
our experience that are invariant and omnipresent, or at least much less
variable than the vast variety of its empirical characteristics.” These are the
“basic categories (with their corresponding concepts) in terms of which we
define men” (Berlin 1980, pp. 165–66). Less ambitiously, the aim may be to
defend, clarify or problematize certain abstract notions that are implicit in
a very broad set of normative theories—for example, abstract notions of law
(Hart 1994) or rights (Hohfeld 2001) or desert (Kagan 2012) or “fitting-
ness” (Cupit 1996) or, as we have seen, freedom. Abstract metatheoretical
reflection of this sort has been somewhat neglected in contemporary politi-
cal philosophy. It does not produce policy prescriptions, but it is no less
important for that. Political philosophy is, after all, a branch of philosophy.
3.3 Concepts and Conceptions

In the literature on freedom it has often been suggested that we should class
as different “conceptions of freedom” the different definitions covered by
MacCallum’s abstract “concept of freedom.” This application of the concept-
conception distinction seems to me to be a mistake.12 Those who endorse
different “conceptions of x” are parties to a substantive disagreement about
the properties grounding the defining elements of a “concept of x”: if x is
defined as “a and b and c,” the parties with the different “conceptions of x”
are in disagreement about what counts as an a and/or what counts as a b and/
or what counts as a c. However, those who adopt different definitions of
freedom rarely appear to be in disagreement (in any relevant sense) about
what counts as an agent, what counts as a constraint, or what counts as
doing or becoming something.
Take, for example, the “constraint” element of MacCallum’s triadic rela-
tion. Steiner holds that a person’s freedom is constrained only when another
person physically prevents her from doing something. The Berlin Wall made
East Germans unfree to emigrate to the West; when the Wall came down,
they became free to do so, even if they had an irrational fear of emigrating or
had been dissuaded from doing so by deceitful government propaganda.
However, Steiner would not say that his position amounts to a view about
what counts as a constraint. Rather, he would say that it amounts to a view
about what counts as unfreedom. Liberals like Steiner hold that only a cer-
tain kind of constraint, a constraint of the humanly imposed, physical kind,
can remove freedom as they understand it, for only this kind of constraint
can remove the freedom-like property the distribution of which is, from
12
A mistake that I too have made in earlier writings. For example: Carter (1999),
pp. 15–17.
302 Ian Carter
their point of view, a concern of justice. They think—for example—that

justice is grounded in the presence or protection of a measure of negative
freedom (distributed in certain ways), but not in the presence or protection
of positive freedom. It seems, then, that negative and positive freedom are
different “concepts” of freedom, not alternative “conceptions” of MacCallum’s
concept. The mere fact that one concept is abstracted from another (as
MacCallum’s is from Steiner’s, among others) does not make the latter a
conception of the former.13
We should use the expression “conceptions of x,” then, only when we
find genuinely substantive disagreement about what counts as one of the
defining elements of a concept of x.14 In the case of freedom, such substan-
tive disagreement seems to occur only at a more specific level than that
identified by MacCallum’s concept. For example, it seems plausible to say
that there are different conceptions of the concept of negative freedom that
depend on genuinely substantive controversies about what counts as being
“humanly imposed.”
A similar line of reasoning applies to the concept of justice. Some con-
cepts of justice occasion controversies concerning the grounding properties
of their defining elements. Others are more abstract and serve to identify a
conceptual structure common to different concepts of justice (they show
why the members of a particular set of concepts are all held to count as
concepts of justice). The “concept of justice” identified by Rawls seems to
me (but not, apparently, to Rawls himself ) to be of the latter kind. Rawls
characterizes “the concept of justice” as involving, among other things, “a
proper balance between competing claims to the advantages of social life”
(Rawls 1999, p. 5). More specific understandings of justice presumably
make specifications about which advantages of social life can be the object
of claims of justice. The differences between these understandings of justice
do not, however, turn on what different people see as counting as “advan-
tages of social life.” Rawlsians, for example, exclude happiness, friendship,
company, and entertainment, from the list of things to which people can
have claims of justice, but we should be surprised to hear them deny that
happiness, friendship, company, or entertainment are “advantages of social
life.” If we disagree about which advantages of social life are relevant to jus-
tice, we are disagreeing about which concept of justice to adopt—that is,
about the definition of justice at a less abstract level.
We justify our use of one concept rather than another by showing how
the former fits into a wider conception of some concept that is located
13
Although MacCallum does not use the concept-conception distinction, he does
seem to think that he has clarified the concept of freedom (MacCallum 1967, p. 320).
14
This seems to have been Hart’s view (1994, pp. 246–7).
higher up in our hierarchy (and, hopefully, in those embraced by others).

Thus, one way of justifying the adoption of a particular concept of freedom is
by showing how it fits into a wider conception of justice. As long as a certain
property, x, is located relatively low down in the grounding hierarchies
of the disagreeing parties—as freedom tends to be—it is relatively easy to
make sense of the uses of different “concepts of x” as reflecting a genuinely
substantive disagreement: we can do so by showing how they reflect differ-
ent conceptions of a shared concept that is located higher up in alternative
hierarchies. As we ascend our hierarchy, it becomes more difficult to find
higher-level concepts the different conceptions of which explain to others
why we prefer one “concept of x” to another.
The identification of an abstract, metatheoretically value-neutral con-
cept is a different matter: it involves abstracting from different concepts.
Here, we encounter another difficulty as we ascend the hierarchy: that of
combining generality and complexity with abstraction. At the higher levels
of generality and complexity it may be difficult to encapsulate an abstract
concept in a way that proves, in Dworkin’s words, “neither so thin as to be
uninformative nor so thick as to fail to be fully inclusive” (Dworkin 1986,
pp. 74–5). It is perhaps this last fact that best explains the greater influ-
ence among political philosophers of MacCallum’s abstract neutral con-
cept of freedom than of any attempt to specify a similarly abstract concept
of justice.
Dworkin rejects the Rawlsian understanding of the distinction between
concepts and conceptions. His preferred model for the identification of a
shared “concept of justice” is that of a shared set of paradigm cases of appli-
cation of the concept. On this model, to disagree substantively about the
validity of a particular “conception of justice” is to disagree about how to
extend applications from agreed paradigm cases of justice and injustice
to more controversial cases. If Dworkin’s way of distinguishing between a
concept and its conceptions were correct, there would be little space for
metatheoretical value-neutrality: the concept of justice, of which the differ-
ent conceptions represent alternative interpretations, would not be suscep-
tible to clear definition (it would be an “open” concept); and there would
certainly be no opportunity to ground a clear abstract concept in a set of
different “concepts of justice.” Even if Dworkin is right about justice, how-
ever, this may only show that the contours of our concepts become more
vague as we ascend the hierarchy. It will not show that we must also apply
his model to mid-level concepts like rights, desert, interests, needs, or
well-being. Certainly, when it comes to capturing our sense that different
theorists of freedom are all “talking about the same thing,” MacCallum’s
concept has proved much more successful than any attempt to point to a set
of paradigm cases.
304 Ian Carter
4. CONCLUSION
The notions of value-freeness and value-neutrality can and do play impor-

tant roles in the clarification of ethical and political ideas. Their importance
depends on the nature of our purposes in engaging in conceptual analysis.
Where our purpose is to construct a coherent normative theory with
practical implications, it matters that some of our concepts are value-free
whereas others are not. In clarifying the concepts employed in such a the-
ory, we need to pay attention to the grounding relations between the prop-
erties to which they refer, where the more general and complex concepts
depend on other more specific and simple ones in order to pick out portions
of the empirical world. By revealing the structure of these grounding rela-
tions, we reveal the order in which we give reasons for describing the world
in certain normatively relevant ways, and for attempting to change it. We
need value-free concepts because our ethical reason-giving must eventually
pass in a manageable way from the ethical realm to the empirical realm it if
is to affect how we view, and participate in, ethical and political events and
states of affairs.
Where our purpose is, more specifically, to construct a normative theory
on which disagreeing parties can converge, so that they can all neverthe-
less endorse a specific set of evaluations and prescriptions, then it will
be important to us to use concepts that display a certain degree of norma-
tive value-neutrality. Such concepts might or might not be value-free in
the above sense.
Finally, where our purpose is the metatheoretical one of clarifying the
ethical categories shared by advocates of contrasting positions or, more
ambitiously, by all humans, we shall be interested in constructing con-
cepts that display a certain degree of metatheoretical value-neutrality. Such
value-neutral concepts are abstractions, and the concepts they are abstracted
from might or might not be value-free. We can use neutral concepts of
this sort to try to assemble general theories either about ethics and human
nature or about certain broad ethical or political traditions. We can also
refer to the internal structure of a neutral concept of this sort in order to
clarify the nature and source of a particular normative disagreement—for
example, the disagreement between advocates of negative and positive
freedom, who are shown to emphasize different dimensions of an abstract,
neutral concept of freedom, or between supporters of rival theories of
rights, who are shown to emphasize different dimensions of an abstract,
neutral concept of rights.
In none of these cases does a concept’s qualifying as value-free or value-
neutral depend on its having the “archimedean” quality of value-independence.
Bibliography
Berlin, Isaiah (1980). Concepts and Categories (Oxford: Oxford University Press).
Berlin, Isaiah (2002). “Two Concepts of Liberty,” in I. Berlin, Liberty (Oxford:
Oxford University Press).
Carter, Ian (1995). “The Independent Value of Freedom,” Ethics, 105: 819–45.
Carter, Ian (1999). A Measure of Freedom (Oxford: Oxford University Press).
Connolly, William E. (1993). The Terms of Political Discourse, third edition (Oxford:
Blackwell).
Correia, Fabrice, and Schnieder, Benjamin (eds) (2012). Metaphysical Grounding.
Understanding the Structure of Reality (Cambridge: Cambridge University Press).
Cupit, Geoffrey (1996). Justice as Fittingness (Oxford: Clarendon Press).
Day, J. P. (1987). Liberty and Justice (London: Croom Helm).
Dworkin, Ronald (1986). Law’s Empire (London: Fontana Press).
Dworkin, Ronald (1996). “Objectivity of Truth: You’d Better Believe It,” Philosophy
and Public Affairs, 25: 87–139.
Dworkin, Ronald (2004). “Hart’s Postscript and the Character of Political
Philosophy,” Oxford Journal of Legal Studies, 24: 1–37.
Dworkin, Ronald (2011). Justice for Hedgehogs (Cambridge, MA: Harvard University
Press).
Hare, R. M. (1984). “Supervenience,” Proceedings of the Aristotelian Society, 58: 1–16.
Hart, H. L. A. (1994). The Concept of Law, second edition (Oxford: Clarendon
Press).
Hees, M. Van (2000). Legal Reductionism and Freedom (Dordrecht: Kluwer).
Hohfeld, Wesley Newcomb (2001). Fundamental Legal Conceptions as Applied in
Judicial Reasoning (Dartmouth: Ashgate).
Jackson, Frank (1998). From Metaphysics to Ethics. A Defence of Conceptual Analysis
(Oxford: Clarendon Press).
Kagan, Shelly (2012). The Geometry of Desert (Oxford: Oxford University Press).
Kovesi, Julius (1967). Moral Notions (London: Routledge and Kegan Paul).
Kramer, Matthew H. (2003). The Quality of Freedom (Oxford: Oxford University
Press).
Kramer, Matthew H. (2009). Moral Realism as a Moral Doctrine (Oxford: Blackwell).
Kramer, Matthew H. (forthcoming). “Conceptual Analysis and Distributive Justice,”
in S. Olsaretti (ed.), The Oxford Handbook of Distributive Justice (Oxford: Oxford
University Press).
Lovett, Frank (2010). A General Theory of Domination and Justice (Oxford: Oxford
University Press).
Lukes, Steven (2005). Power. A Radical View, second edition (London: Macmillan).
MacCallum, Gerald C. Jr (1967). “Negative and Positive Freedom,” The Philosophical
Review, 76: 312–34.
McLaughlin, Brian, and Bennett, Karen (2011). “Supervenience,” in Stanford
Encyclopedia of Philosophy, <http://plato.stanford.edu/entries/supervenience>.
Morriss, Peter (2002). Power. A Philosophical Analysis, second edition (Manchester:
Manchester University Press).
306 Ian Carter
Oppenheim, Felix E. (1961). Dimensions of Freedom. An Analysis (New York, NY:

St. Martin’s Press).
Oppenheim, Felix E. (1981). Political Concepts. A Reconstruction (Chicago: Chicago
University Press).
Oppenheim, Felix E. (1995). “Social Freedom and its Parameters,” Journal of
Theoretical Politics, 7: 403–20.
Putnam, Hilary (1981). Reason, Truth and History (Cambridge: Cambridge
University Press).
Rawls, John (1999). A Theory of Justice, revised edition (Cambridge, MA: Harvard
University Press).
Ricciardi, Mario (2007). “Berlin on Liberty,” in G. Crowder and H. Hardy (eds),
The One and the Many. Reading Isaiah Berlin (New York, NY: Prometheus).
Sheffler, Samuel (1987). “Morality Through Thick and Thin. A Critical Notice of
Ethics and the Limits of Philosophy,” The Philosophical Review, 46: 411–34.
Smith, Michael (2000). “Does the Evaluative Supervene on the Natural?,” in
R. Crisp and B. Hooker (eds), Well-Being and Morality. Essays in Honour of James
Griffin (Oxford: Clarendon Press).
Stalnaker, Robert (1996). “Varieties of Supervenience,” Philosophical Perspectives, 10:
221–41.
Steiner, Hillel (1994). An Essay on Rights (Oxford: Blackwell).
Strawson, P. F. (1992). Analysis and Metaphysics (Oxford: Oxford University Press).
Sugden, Robert (2003). “Opportunity as a Space for Individuality: Its Value and the
Impossibility of Measuring It,” Ethics, 113: 783–809.
Taylor, Charles (1979). “What's Wrong with Negative Liberty,” in A. Ryan (ed.), The
Idea of Freedom (Oxford: Oxford University Press).
Williams, Bernard (2001). “From Freedom to Liberty: the Construction of a Political
Value,” Philosophy and Public Affairs, 30: 3–26.
Index
References to footnotes are given in italic type.
abstract rights 200, 218–20 common law 251, 252

Acadians 156 communication 17–18
adultery 242 compensation (for intrusions) 204–7
advantage-taking see exploitation concept analysis, ethical judgements 279
agency 48 concepts and conceptions 301–3
bending 54, 57–8 Conditional Force Argument 173–4
in voting 47–9, 52–5 connective analysis 290
aggregation 187–8 Connolly, William 282–5, 296
Altman, Andrew 154–8 consent 127
annexation 165–6, 168, 170 consequentialism see rights
Aquinas, Thomas 25, 254, 257 consequentialism
Archimedianism 283, 286, 295–6 Constant, Benjamin 67
Aristotle 254 cosmopolitanism 249
ash deposition 211–12 courts 28, 238
autonomy 109, 131–3, 139, 205
decision-making 13–16
Bamford v Turnley 201, 208–11, 218 see also voting
Barry, Brian 94, 102, 109 democracy 27–8
basic (social) structures 9–10, 11, 102 democratic approval 1, 16–22
justice 10 democratic control 1, 22–30
Beitz, Charles 11 people as corporate agency 29
Berlin, Isaiah 23, 71–2, 301 deontological theories of justice 221, 224–5,
on freedom 299, 301 231–5, 236, 241, 244–5, 248
Berlin Wall 301 Digest of Roman Law 25
bloodless invasions 174–5 disagreement 68–74
Bodin, Jean 22, 65 displaced persons 166–7
Bomb 183 Dreben, Burton 125
bracketing strategy 88–92, 93, 96, 98, duties
101, 109 agent-neutral 230
Bramwell, George 201, 208–11, 216 of care 183–6
bullying 177 Dworkin, Ronald 13, 282, 283–6, 291,
295–6, 300, 303
Carroll, Lewis 21
Catholicism 114, 115, 117, 130–1, 138 egalitarianism 13
causation (in voting) 47–9 elbow room postulate 197–8, 212–17
choice deprivation 202 elections 1–2
Christianity 64 British Parliamentary 40, 188
see also Catholicism German 43
Christiano, Thomas 11 United States Presidential 37, 39–40
cities, liveability ranking 9 see also democracy; voting
classical liberalism 283 Epstein, Richard 195, 200–1
coercion 130–4 in-kind compensation 198, 204–7,
Cohen, G.A. 26, 199, 235, 243 208, 210
308 Index
Estlund, David 114, 138 Goldman, Alvin 38–9, 41–4, 52–8

ethical evaluation 279–80 divergence from Goldman 45–9
evaluative concepts 282, 284, 289 Great Britain 156
exploitation 5, 250 parliamentary elections 40, 188
benefit at others’ expense 262–3 Grotius, Walter 22
common law account 251 grounding properties 290–4
distinctive character of ethical in empirical 292
wrongfulness 273
doctors 258–9 Habermas, Jürgen 11, 19, 82
examples Hampton, Jean 122
diminished capacity 271–2 Hart, H.L.A. 203, 283
rescue 265–7 Hayek, Frederic 28
unequal opportunity 267–71 Hitler, Adolf 43, 53
fairness and 272–3 Hobbes, Thomas 22, 25, 65
hybrid accounts 258–60, 260–1 hog-tying 5, 195
Marxist account 251 hostile takeovers 174, 188–9, 190–1
procedural accounts 252–4, 260–2 hypotheticalization, consent 127–8
prostitution 255
psychotherapy 257 idealization 119
substantive accounts 254–8, 260–2 individualism, war and, Conditional
sweatshop labour 257, 264 Force Argument 174–6
tow truck operators 253–9 individuality 66
violation of duties 263 Islamic fundamentalism 280
as wrong 262–4
eyeball redistribution 204 Jews 26
eyeball test 27, 29 Julius, A.J. 266
jurisdictional rights 161
fairness 203, 223, 239–41 just war theory 173
exploitation and 272–3 justice 288, 298
see also exploitation concepts and conceptions 32
Hart’s principle 203 deontological theories 221, 224–5,
justice as 102, 223, 239 231–5, 236, 241, 244–5, 248
Farrelly, Colin 223 fairness and 102, 223, 239
Filmer, Robert 213 freedom and 290
filming 262–3 moral conception 102
forced exchanges 204 political 1
Forst, Rainer 11 basic 30–1
freedom 23–4, 288 conditions for 32
conceptions of 301 conditions for a state to be 10–11
grounding justice 290 conformity to moral justice 102–8
loss through non-compliance by definition 10
others 234 democratic decision-making
non-specific concept 291 and 12–15
functionalist territorial rights elaborate and basic theories 16–17,
account 148–9, 150–2, 170–1 29, 30
boundary problem 153 justice
social and 11–12, 32
Gans, Chaim 147, 150 priority over 15–16
Gaus, Gerald 85, 116, 125 practical implications 32–3
genocidal aggression 174, 176, 180–2, primacy over justice, social 31–2
184, 186 stability 91–2
geodesic domes 296 private 24
globalization 86–7 public 24–5
Index 309
robust claim-satisfaction 24–5 political independence problem 93–6,

social 1, 26–7 101–2
constraining lawgivers and balancing 99
policymakers 26 dualist solution 99–100
decision-making and 14–21 moral action 97–8
definition 10 political action 96–7
democratic control and 22–30 principle of legitimacy 75
disposition of forbearance 12–14, reasonable disagreement 68–74
22–3 societal characteristics 65
egalitarianism 13 libertarian rights 201–2
normative theory 15, 23 hog-tying objection 195–7
political justice and 11–12, 32 liberty 300
priority of political justice over 12, lifeguards 39
31–2 liveability ranking 9
recommendations 17–19 live-and-let-live rule 208–11
status of philosophical discussion 30 Locke, John 65, 213
stability of theories 90, 96 on individual rights 202
on joint-ownership of nature 213, 215
Kant, Immanuel 65, 78 on private property 213–14
reciprocity 124 on states’ territorial rights 158–63
states’ rights 169, 170 Lockean rights theory 198–201
kidnapping 243 liability rule attenuation 201–4
kidney redistribution 204 see also libertarian rights
killing, individual rights 175–6 London 188
Kovesi, Julius 292, 293 Lovett, Frank 282
Kramer, Matthew 282, 294 Lukes, Steven 282
landowners 161 MacCallum, Gerald 299–303

Larmore, Charles 26 Mackie, J.L. 41
legal orders 164 malicious intrusions 211–12
legitimacy (political) 75, 77, 126–9 market prices 255
accessibility under public-reason markets 37
theories 114–15 Marx, Karl 122, 251, 254, 255
basis in respect 75–6 Maryland v Shatzer 243
hypothetical justification 126–9 mediated harms 183–6
liberal principle of 75–7, 83 Mill, John Stuart 65, 122
lesser interests 187–8 Miller, David 150–2, 236
levelling down 272–3 minor intrusions 197–8, 201
liability rule compensation 201–4 compensation 204–7
liability rules 196, 196–7, 204, 216–17 conditions governing 218
liberalism 2–3 elbow-room argument 213–16
abstract principles 65 implicit in-kind compensation 204–7
application to global problems 86–7 live-and-let-live rule 208–11
bracketing strategy 88–92, 96, 101–2 Montaigne, Michel de 65
classical 66 Montevideo Convention (1933) 145
contemporary relevance 86 Moore, G.E. 18
development 64–5 moral counterproductivity 253
exclusion 85, 85–6 moral obligation
external conception 80 political obligation and
inclusiveness 84–5 balancing 99
individualism 66 dualism 99–101
latecomer status 65–6 primacy of moral obligation 97–8
novelty 63–4, 65 primacy of political obligation 96–7
310 Index
moral reasons 92–3 non-architectonic 103–8

moral responsibility (voting) 45–9 political justice 1
moral rules 77 basic 30–1
moral self-satisfaction 52 conditions for 32
Morriss, Peter 288 conditions for a state to be 10–11
conformity to moral justice
Nagel, Thomas 13, 116, 136 architectonic argument 102–3
nationalism 148–9, 150–2, 154 non-architectonic argument 103–8
hybrid accounts 165 definition 10
natural resources 149 democratic decision-making and
Nazi, see also Hitler, Adolf 12–15
Nazi Germany 26, 27 elaborate and basic theories 16–17,
Nine, Cara 146 29, 30
non-compliance 222 practical implications 32–3
costs 240 primacy over social justice 31–2
deontological theories 222–3, 231–5 social justice and 11–12, 32
remedial approach 223–4, 226–9 priority over 15–16
rights consequentialism 224, 229–31 stability 91–2
rights-generating rules 234 political liberalism 2–3
sources 232 abstract principles 65
Nozick, Robert 194–7, 200, 201–4, application to global problems 86–7
205, 214 bracketing strategy 88–92, 93, 96, 98,
Nussbaum, Martha 72 101, 109
contemporary relevance 86
Obama, Barack 280 development 64–5
obligation exclusion 85–6
political external conception 80
moral obligation and 96–8 inclusiveness 84–5
balancing 99 individualism 66
dualism 99–100 latecomer status 65–6
Oppenheim, Felix 282, 284–5, 287, 297, novelty 63–4, 65
297–8, 299–300 political independence problem 93–6,
Otsuka, Michael 245 101–2
balancing 99
papal infallibility 138 dualist solution 99–100
permission 95 moral action 97–8
personhood 77–8 political action 96–7
physical invasion test 207 principle of legitimacy 75
Pinch 177, 179 reasonable disagreement 68–74
pivotal votes 36 societal characteristics 65
causal responsibility and 38–9 political liberties 11
pleasure 23 political obligation
pluralism 71–2, 91, 165–71 moral obligation and
wrongs of displacement 166–7 balancing 99
Pogge, Thomas 261, 267 dualism 99–100
police officers 39 primacy of moral obligation
political action 97 97–8
political aggression 175–6 primacy of political obligation
political association 63–4 96–7
political attitudes 53 politics (practice of ) 134–7
political independence 93–4, 101–2 Pope 114
arguments against power, legitimacy of see legitimacy
architectonic 102–3 power relations 279
Index 311
pragmatic inconsistency 18 Raz, Joseph 79, 117, 122, 133, 136

price gouging 253 reasonable disagreement 68–74
principle anti-paternalism 202–3 reasons 91–2
private justice 24 for action 131–4
private property see property, private moral 92–3
Proliferation Problem 174, 188–9 of respect 105–6, 106–8
property 167–8 recommendations 17–19
private 213–14 reduced capacity 270–2
value 254–5 reductive individualism
property rules 196 Conditional Force Argument 174–6
proportionate defense 183 genocidal aggression 180–2
prostitution 255 proliferation problem 188–9
prudence 94 rights and communities 190–2
psychotherapists 257 vital interests 178–80
public justice 24–5 regress argument 157
public reason 3, 112–13 religion 64, 84, 242
hypothetical conditions 126–30 see also Catholicism
idealization rescue 265–7
acceptability of 118–20 respect 76–7, 79
necessity of 117–18 responsibility 236–9
reasons for action 130–1 Ridge, Michael 240
reciprocity principles and 123–4 right-generating rules 234
role of political philosophy 134–7 rights
underlying intuitions 114–16 abstract 200
unreasonableness and 121–2 defensive 190–2
exclusion 124–5 minor intrusions 197–8, 201
Putnam, Hillary 284 implicit in-kind compensation
204–7
Quong, Jonathan 80, 91, 121, 125, 134 live-and-let-live rule 208–11
wanton and malicious 211–12
racism 245–6, 247 private property 213–14
Railton, Peter 194 rules and 218–20
rape 178–9, 182 self-ownership see self-ownership rights
Rational Choice Theory 36 of self-ownership 4, 195–7, 201–2
pivotal votes 36 soft paternalism 217
causal efficacy of voting and 38–9 territorial see territorial rights
voting responsibility 44–5 rights consequentialism 224, 229–31, 233
rationality, elections 1–2 departure from ideal theory sacrifice
Rawls, John 2, 11, 74, 99–100, 302 justice 235–6
account of states 148 fairness objection 239–41
on distinction between political impermissibility of infringement of
philosophy and politics 134 moral rights 244–5
on exclusion of unreasonable 124–5 moral rules justifying moral
on just basic structures 11 rights 245–6
liberal principle of legitimacy responsibility objection 236–9
75–7, 83 robbers 39
on pluralism 71–2 Rodin, David 4
on political justice 102 on war morality 173–4
stability 90–1, 100 duties of care 183–6
on political liberties 11 genocidal aggression 180–2
on primacy of political values 98 mediated harms 176–80
reasonableness 120 vital interests 178–80
on respect 79 Rousseau, Jean-Jacques 22, 29, 85
312 Index
rules Stemplowska, Zofia 241

justifying rights 245–6 Stilz, Anna 148, 160, 165–71
non-compliance, responsibility Strawson, P.F. 290
objection 236 Stroud, Sarah 95
not determining moral rights 242–3 Stroud, Sarah 95
rights and 218–20 Sugden, Robert 294
rights consequentialism 236–41 sweatshop labour 253, 264, 268–9
Syria 135
Schmidtz, David 202
Schmitt, Carl 85 territorial rights
self-defense 175, 183, 214 competing claims 151
self-ownership rights 4, 194–5 functionalist account 148, 152–4
Sen, Amartya 21, 230 individualist voluntary account
Sheffler, Samuel 288 158–63
slavery 204 majoritarian voluntarist account 154–8
Sobel, David 194–5 nationalist accounts 148–9, 150–2
social justice 1, 26–7 other accounts 163–4
constraining lawgivers and pluralist account 165–71
policymakers 26 Stilz’ account 165–71
decision-making and 14–15, 15–16 voluntarist account 147–8, 149
democratic approval 17 tools 203
multiple processes 19–21 trapped minorities 155
definition 10 Troy 105–6
democratic control and 22–30 Tuck, Richard 38–41
disposition of forbearance 12–14, 22–3
egalitarianism 13 unequal opportunity 267–71
normative theory 15, 23 unfair advantage 261–2
political justice and 11–12, 32 United States 135
priority of political justice over 12, 31–2 Mexican border 153
recommendations 17–19 Presidential elections 37, 39–40
status of philosophical discussion 30 unreasonableness 121–2
soft paternalism 217 methodological exclusion 124
states
annexation 165–6, 168, 170 Valdman, Mikhail 266
definition 145 Valentini, Laura 12
jurisdictional claims 146 value (of property) 254–5
cosmopolitan account 149 value-freeness 5–6, 280, 287
Kantian account 148 description and evaluation 287–9
natural resources 149 necessity of 294–6
rights claimed 145–6 value-independence 285–6
territorial rights value-neutrality 5–6, 280, 282–3, 285,
competing claims 151 296–7
functionalist account 148, 152–4 metatheoretical 298–301
individualist voluntary account normative 297–8
158–63 utility 297
majoritarian voluntarist values, idealization 119–20
account 154–8 vista alteration 207
nationalist accounts 148–9, 150–2 voluntarism 148–9, 247–8
other accounts 163–4 hybrid accounts 165
pluralist account 165–71 individualist 158–63
voluntarist account 147–8, 149 majoritarian 154–8
Steiner, Hillel 228, 282, 294, 300, voting 20–1
301–2, 304 abstention 56–7
Index 313
agency effects 47–9, 55–6 Waldron, Jeremy 14, 114, 167, 229
action-guiding advice 49–52 Wall, Steven 89–90
choice and 48 war
causal efficacy of 40–1 collectivist accounts 173
Goldman’s account 40–1 Conditional Force Argument 174–6
sufficient conditions for reductivist account 173
outcome 42 duties of care 183–6
Tuck’s account 38–41 genocidal aggression 180–2
causal responsibility for outcome 43, lesser interests 187–8
43–4, 57 mediated harms 176–8
expressive account 53–5 proliferation problem 188–9
perceived 56–7 rescue cases 182–3
expressive account 37 vital interests 178–9
electoral responsibility 53–5 Welfare as Reason 92–3, 101, 104
moral responsibility 45–9 Wellman, Christopher Heath 154–8
for outcome 54 Wertheimer, Alan 254, 257, 261
pivotal votes 36–7 Williams, Bernard 293
political attitudes and 53–4 Wood, Allen 252
rational choice theory 36–45 wrongful exploitation 5
rational ignorance 37 wrongful subjection 169–71
rationality of 1–2, 37
Vrousalis, Nicholas 266 Zwolinski, Matt 261

David Sobel, Peter Vallentyne, Steven Wall - Oxford Studies in Political Philosophy, Volume 1-Oxford University Press (2015) PDF

Încărcat de

Informații document

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

David Sobel, Peter Vallentyne, Steven Wall - Oxford Studies in Political Philosophy, Volume 1-Oxford University Press (2015) PDF

Încărcat de

Drepturi de autor:

Formate disponibile

OUP CORRECTED PROOF – FINAL, 25/03/15, SPi

OXFORD STUDIES IN POLITICAL PHILOSOPHY

Most of the chapters assembled here were first presented at a workshop in

Part I: Democracy

Part II: Political Liberalism and Public Reason

Part III: Rights and Duties

Part IV: Method

Geoffrey Brennan is Distinguished Research Professor in the Department

With this inaugural volume, we are pleased to introduce Oxford Studies in

Standard rational choice theory gives a negative answer. The probability

While potentially controversial, the authority of the moral principle of

despite their disagreements. Ultimately, Carter concludes, it is the failure to

1. THE ISSUE

* I benefited enormously from the discussion of versions of this chapter presented

Justice: Social and Political 11

Whether a given state is politically just depends, under a broad conception,

Justice: Social and Political 13

2. SOME POLITICALLY JUST PROCESS OF

likely to be many different views on what social justice requires. Whatever

Justice: Social and Political 15

is committed, as we might put it, to the idea that while it is theoretically

3. THIS PROCESS OUGHT TO REQUIRE A FORM

Justice: Social and Political 17

ought to be established in any society for which they make prescriptions.

Justice: Social and Political 19

The requirement introduced may seem to be excessively demanding,

Justice: Social and Political 21

A salient candidate may emerge with someone’s making a proposal to

hold, rightly or wrongly, that such an arrangement promises a good prospect

4. THIS PROCESS OUGHT TO INTRODUCE A FORM

The forbearing character of theories of social justice—and elaborate theories

Justice: Social and Political 23

society there ought to be a just process for accommodating differences between

Justice: Social and Political 25

that the institutions of the society ensure your claim-satisfaction actually

social protection or insurance in the event of suffering penury, unemploy-

Justice: Social and Political 27

compromise claims it supports. The process should be sufficient to ensure

Justice: Social and Political 29

even if the governing agency happened to grant people’s claims in social

5. THEORETICAL AND PRACTICAL IMPLICATIONS

There are two theoretical implications of the argument in this chapter,

Justice: Social and Political 31

But whereas we put forward proposals in social justice—and in the elab-

equal voices: they are equally positioned to be heard—perhaps via repre-

Justice: Social and Political 33

it would amount to trying to privilege the philosophical seminar over the

Forst, R. (2002). Contexts of Justice: Political Philosophy beyond Liberalism and

Justice: Social and Political 35

In the standard Rational Choice Theory (RCV) account of voting behavior,

Voting and Causal Responsibility 37

38 Geoffrey Brennan and Geoffrey Sayre-McCord

An implication of RCV is that any voter who votes in standard large-­

Voting and Causal Responsibility 39

40 Geoffrey Brennan and Geoffrey Sayre-McCord

Voting and Causal Responsibility 41

42 Geoffrey Brennan and Geoffrey Sayre-McCord

Voting and Causal Responsibility 43

44 Geoffrey Brennan and Geoffrey Sayre-McCord

Voting and Causal Responsibility 45

3. We should expect that considerations of individual self-interest

2. RESPONSIBILITY FOR WHAT?

46 Geoffrey Brennan and Geoffrey Sayre-McCord

An implication of RCV is that any voter who votes in standard large-